Population Health in the Dutch Caribbean

Population Health in the Dutch Caribbean

A comparative study of political context and

health policy performance

Soraya P.A. Verstraeten

Colofon

ISBN: 978-94-6050-031-2

Population health in the Dutch Caribbean.

A comparative study of political context and health policy performance.

Doctoral thesis, Erasmus University Rotterdam, the Netherlands

© Soraya P.A. Verstraeten, 2020

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any

form or by any means, without permission from the author or, when appropriate, from the copy-

right holding publishers of the publications.

The research presented in this thesis was performed at the Institute of Public Health Curaçao

(Volksgezondheid Instituut Curaçao, VIC), Ministry of Health, Environment and Nature,

Willemstad, Curaçao & the Department of Public Health, Erasmus MC, Rotterdam, the

Netherlands. These studies used data from the National Health Surveys Curaçao and the

Netherlands, and from open access data sources, such as the World Health Organization’s

mortality database, the Pan American Health Organization’s core health indicator database and

the United Nation’s World Populations Prospects database. The National Health Survey Curaçao

2013 was conducted and funded by the Institute of Public Health (VIC). The Dutch Public Health

Monitor 2012 of the Netherlands was conducted and funded by the National Institute for Public

Health and the Environment (RIVM).

This thesis was printed with the financial support of the Ministry of Health, Environment and

Nature, Willemstad, Curaçao.

Cover design and print: One Media Group, Willemstad, Curaçao

Population Health in the Dutch Caribbean A comparative study of political context and

health policy performance

Volksgezondheid in de Nederlandse Cariben Een vergelijkende studie van de politieke context en

prestaties van het gezondheidsbeleid

Proefschrift

ter verkrijging van de graad van doctor aan de

Erasmus Universiteit Rotterdam

op gezag van de

rector magnificus

Prof.dr. R.C.M.E. Engels

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

woensdag 8 april 2020 om 13:30 uur

door

Soraya Petronella Adriana Verstraeten

geboren te Breda

Promotiecommissie

Promotoren Prof.dr. J.P. Mackenbach

Prof.dr. J.A.M. Van Oers

Overige leden Prof.dr. P.J.E. Bindels

Prof.dr. S. Denkta

Prof.dr. A.E. Kunst

Contents

Chapter 1 General introduction 6

Part I - Population health comparisons within the Kingdom of the Netherlands

Chapter 2 The health of Antillean migrants in the Netherlands: a comparison with the health of non-migrants in both the countries of origin and destination

28

Chapter 3 Differences in amenable mortality between the constituent countries of the Kingdom of the Netherlands

50

Part II - The political context of health and health policy performance in the Caribbean

Chapter 4 Decolonization and life expectancy in the Caribbean 68

Chapter 5 Differences in life expectancy between four Western countries and their Caribbean dependencies, 1980-2014

92

Chapter 6 Health policy performance in 16 Caribbean states, 2010-2015 110

Chapter 7 General discussion 128

English summary 161

Nederlandse samenvatting 169

Resúmen na Papiamentu 177

Acknowledgements 185

About the author 187

List of publications 189

PhD portfolio 191

Chapter 1

General introduction

Chapter 1

8

The Netherlands is at the forefront of population health monitoring (1), but not much is known

about the health of inhabitants in the Caribbean territories of the Kingdom of the Netherlands.

The available information seems to fit Sir Michael Marmot’s statement that “all too commonly

where health is poorest, health information tends to be poorest” (2). Life expectancies are shorter,

and infant and maternal mortality rates are higher, in the two largest Dutch Caribbean islands,

Aruba and Curaçao, than in the Netherlands (3,4). Self-reported obesity is twice as prevalent in

Curaçao than in the Netherlands, and goes hand-in-hand with increased rates in obesity-related

conditions such as diabetes mellitus, hypertension and musculoskeletal disorders (5). Obesity,

diabetes mellitus and hypertension are more prevalent among Dutch Caribbean migrants living in

the Netherlands as well (6–8). In addition, the risk of mortality from diabetes is three to four times

higher among Dutch Caribbean migrants in the Netherlands, compared to the Dutch population

(9).

The causes of the poorer health outcomes of the Dutch Caribbean people are largely unknown,

but are commonly ascribed to unspecified regional, biological or cultural characteristics. Beyond

these general descriptions, however, more specific explanations are necessary to inform policy

strategies that are aimed at improving the health of inhabitants in the Dutch Caribbean islands.

The urgency for this has not gone unrecognized by the governments of Aruba, Curaçao and St.

Maarten, who have aligned their current governing programs to the 2030 UN Sustainable

Development Agenda (10–12). As the successful implementation of the proposed policies is

fundamentally dependent on the political decision-making process, this thesis aims to provide a

better insight in the ways in which political determinants are associated with health outcomes in

the Dutch Caribbean.

The Dutch Caribbean

The political context

Apart from several short interruptions, the Dutch Caribbean islands are governed by the Dutch

since the seventeenth century. The two largest islands, Aruba and Curaçao, were both territories

of the former Netherlands Antilles, the autonomous successor of the Dutch colony Curaçao and

its dependencies (in Dutch: Kolonie Curaçao and onderhorigheden). In addition to these islands,

it consisted of the territories of Bonaire, Saba, and St. Eustatius and St. Maarten. Established in

1954, the Netherlands Antilles has since then fragmented into smaller constitutional parts within

the Kingdom of the Netherlands. Aruba seceded in 1986 as a constituent country within the

Kingdom of the Netherlands, which paved the way for recurring discussions on the constitutional


9

arrangements of the remaining Dutch Caribbean islands. These discussions culminated in the

dissolution of the Netherlands Antilles on 10 October 2010, which was remodeled into two

additional constituent countries, Curaçao and St. Maarten, and the integration of Bonaire, St.

Eustatius and Saba into the Netherlands as special municipalities (see figure 1).

As health data on the other Dutch Caribbean islands are even scarcer, Aruba and Curaçao are the

territories of main focus in this thesis. The islands are located in the southern Caribbean Sea off

the north coast of Venezuela and have the largest populations of the Dutch Caribbean territories.

Nevertheless, their populations are considered small even within the global family of small island

states, with the largest, Curaçao, counting approximately 158.000 inhabitants and Aruba,

approximately 104.000 inhabitants in 2015 (3). The islands have their own constitution, legal

system and democratically elected government. The island’s governments have primary

responsibility for the provision of healthcare and public services for their inhabitants, for

governing the social, educational and health sector, and for steering and evaluating local policy

initiatives. For matters of military defense and foreign policy, the islands are dependent on the

Netherlands.

Figure 1 Constitutional organization of the Kingdom of the Netherlands, from 1954-2010 and from 2010 onwards

Chapter 1

10

The Kingdom council (in Dutch: Rijksministerraad) is the institutionalized guarantor of legal

certainty, good governance and human rights in the Kingdom (13), and the four constituent

countries participate as equal partners. In practice, the Netherlands has the majority vote in the

Kingdom council as it comprises approximately 98% of the Kingdom’s population and territory.

Recurrent topics of conflict between the governments of the constituent countries in the Kingdom

involve the scope of the Council’s authority, and on its limits to hold other countries accountable

when responsibilities are not met. In addition to these political disputes, the shared history of the

transatlantic slave trade, slavery and colonization creates recurring societal tensions to the present-

day. Conversely, this shared history also provides the inhabitants of the Dutch Caribbean islands

with strong social, economic, cultural and familial ties to the European part of the Kingdom.

The regional context

Health in the Caribbean region

The Caribbean region is more advanced in its epidemiological and demographic transition relative

to the African region, Asia and Oceana, based on a higher life expectancy at birth and a lower

percentage of deaths due to communicable diseases (14). Because of this, the region is now faced

with the growing burden of non-communicable disease among its aging populations. The

populations of Caribbean states suffer from high rates of non-communicable disorders (15),

worsened by high rates of obesity (16). That this burden has risen to epidemic proportions first

became evident in the 2006 report of the Caribbean Commission on Health and Development

(17). However, to date, an effective response from regional health organizations and local

governments has been lacking (18). Moreover, the region faces the highest mortality rates for

interpersonal violence in the world (19). In 2017, almost 5% of all deaths in Latin America and the

Caribbean were attributed to homicide (20). Other prominent health challenges the Caribbean

states face are HIV/AIDS, mental health disorders and the strengthening of health systems and

the public health infrastructure (17).

Recent publications acknowledge that Caribbean states face specific challenges to strengthen their

health systems and to take efficient action against avoidable mortality (21,22). For starters, out of

31 Caribbean states, 27 are islands or archipelagoes, of which 10 have populations smaller than

100.000. In addition to small size and geographic remoteness, shared characteristics also include

fragile economies, ethnolinguistic diversity, large socio-economic inequalities, emigration of skilled

workforce (brain drain), vulnerability to natural disasters, and a history of colonialism. This shows

that the governments’ capacity to respond to its population needs, as well as global challenges as


11

economic growth, climate change, and organized crime, cannot be done with the same level of

human capital and financial resources that is the norm in larger countries.

Despite these similarities among Caribbean states, the Caribbean is one of the most diverse regions

in the world. As diverse as the Caribbean populations are in terms of their ethnic backgrounds,

cultural identities and religion, as diverse are their health outcomes. With the exception of Canada,

the highest (Martinique, 82), and the lowest (Haiti, 64) life expectancies in the Americas are found

in the Caribbean islands (23).

Health in the Dutch Caribbean

A regional comparison of life expectancy data shows that from the 1950s until the 1990s, some of

the best population health outcomes in the Caribbean region were found in Aruba and Curaçao.

During that period, life expectancy at birth increased from 58.4 years to 73.5 years in Aruba and

from 58.8 years to 74.6 years in Curaçao (3). Although curative care remains the primary focus of

most Caribbean governments (24), including on Aruba and Curaçao, these stunning advances in

longevity may be less the result of medical care services (25,26), and more of improvements in

sanitation and nutrition (27), and the implementation of policies that improved social conditions

and the physical environment (26).

Since the mid-1990s, however, Aruba and Curaçao have experienced a slowdown, and for several

years even a decrease, in life expectancy growth relative to other Caribbean states (3). As a result,

the islands dropped in their rank of the Caribbean life expectancy rating between 1960 and 2015,

from 3rd place to 9th place for Aruba, and from 2nd place to 6th place for Curaçao (out of 21

Caribbean states). So far, there has not been a compelling explanation for this observation. Is there

a Caribbean counterpart that performs arguably better? And if so, what are they doing that the

Dutch Caribbean islands are not? Since other Caribbean states face similar challenges in improving

and protecting the health of their inhabitants as the Dutch Caribbean islands, the ability to

understand health inequalities between Caribbean states would not only be of interest to Dutch-

Caribbean policy makers, but also has become a key objective of public health actions across the

region (24,28).

Social and political determinants and population health

There is a long standing tradition of the idea that political determinants need to be brought into

our understanding of health inequalities within, and between, countries. This idea can be traced

back to the mid-19th century, when pioneers of public health such as Rudolf Virchow, Friedrich

Chapter 1

12

Engels and Edwin Chadwick identified poverty, ignorance from a lack of educational opportunities

and squalid and overcrowded housing as the origins of poor health and recurring epidemics among

peasants and manual workers in Europe (29–31). In order to improve the conditions in which the

working classes lived and worked, they advocated for the necessity of social reforms. The

successful implementation of these reforms, however, meant that decision-makers and -influencers

in society needed to agree on the redistribution of resources and on the collective provision of

social services, which was only possible through continuous political commitment. It is in this

light, that Rudolf Virchow coined his well-known statement: "Medicine is a social science, and

politics is nothing else but medicine on a large scale" (29).

Social determinants of health

The viewpoints of Virchow, Engels and Chadwick still hold significant influence in public health

today, as evidenced by a rapidly increasing number of empirical studies on the role of politics on

population health (32), and a renewed interest in the social determinants of health (33–35). The

paradigm of social determinants poses that health cannot be understood from investigating

individual-level factors such as biology and health behaviors alone, but needs to be considered in

the light of the physical and social environment in which these biological interactions and

behaviors take place. The pathways through which individual social determinants influence

population health are well understood (36). The unequal distribution of social and economic

factors as income, employment, education and housing, for example, produces inequities in health

(37). Even the effect of genes on health, except for several well described exceptions, cannot be

considered separately from the social context of an individual (38).

One of the most widely used ecological models to conceptualize the complex interactions of social

determinants on the health of individuals is the one by Dahlgren and Whitehead (figure 2). The

model depicts the influences on an individual’s health as a dynamic, multifaceted system that

connects upstream determinants as environmental and cultural factors with more proximal ones

like education and housing, and eventually to the individual determinants of biology and health

behaviors. The model was originally developed to describe the structural drivers of health

disparities within countries, but also shows how health may differ between countries as the

structural drivers differ between countries as well. For example, the consumption of healthy foods

has a strong social patterning in different European countries, but could also explain the variations

in health inequalities between these countries (39). By definition, improvements in social

determinants mostly lie outside the influence of medical care practitioners and in the arena of

political action, policy interventions and public services. This does not refute the importance of


13

health care services on health outcomes, but rather shows that an important political and social

dimension underlies who becomes sick or injured in the first place (40–42) and that therefore,

health considerations need to be integrated into policymaking across sectors to improve the health

of all communities in a country (43).

Political determinants of health

When key objectives have been identified, political decision-making will decide which policy

interventions are implemented, what resources will become available, and how different

stakeholders are aligned together and agree on common actions to reach a mutual goal. The

decisions on prioritization and strategy are preferably based on the best available research evidence,

but not enough information may be available for decision-makers to reach more substantiated

conclusions. Examples of health policy interventions that have been proven effective in advancing

population health efforts are improved access to contraceptives to reduce teenage pregnancies

(44), comprehensive bans on the advertisement of tobacco (45) and alcohol (46) to reduce

consumption, infrastructural changes to road design to ensure pedestrian safety and speed control

(47), the implementation of a sugar tax to reduce the consumption of sweetened beverages (48),

breast- and cervix cancer screening for the early detection of cancer (49,50), and the incremental

introduction of a wide range of medical interventions to reduce deaths from perinatal causes (51).

Other health policies aim to strengthen the healthcare system, for example the implementation of

legislation and protocols to ensure the provision and quality standards of healthcare services. Some

policy interventions have an indirect beneficial effect on population health through the reduction

Figure 2 The Dahlgren and Whitehead model of social determinants of health

Chapter 1

14

of socio-economic inequalities, for example welfare state policies (52,53), and policies that address

income inequalities (54,55). What is less clear, however, is whether and why countries vary in their

pursuit of policy interventions, and why some governments are more committed to improve the

health of their population than others.

The political dimension of population health is obvious during times of crisis, for example in the

political back-and-forth on the Affordable Care Act (Obamacare) and in reports of the human

death toll and population displacements during wars. In many other cases, however, the

relationships between political decisions and health outcomes are less obvious. Yet, political

decision-making is at the heart of governmental actions, and inactions, to improve population

health (40–42). Whether or not a certain topic makes it to the policy agenda, and what actions will

be subsequently taken, is dependent on the political perceptions of the severity of a health problem,

the actors who are responsible and the populations who are affected (56). Before discussing the

empirical literature of political determinants on population health outcomes, we first define the

meaning of the word “politics” as the “practice of the art or science of directing and administrating

states” (57). In more concrete terms, politics is concerned with the way that people living in groups

make decisions (58).

As political views, systems and ideologies substantially differ between countries, political

determinants are most commonly conceptualized in four key features: democracy, welfare state,

political tradition, and globalization (32). Democracy reflects the extent of free and fair elections

that is allowed by political regimes (59) and is thought to benefit population health through

pressures on government accountability and responsiveness (60,61). Many studies have focused

on the relation between democratization and health, in particular by looking at the health impact

of the transition from an autocratic to a democratic government (62–68), the association of

harmonized indexes such as Polity IV (69,70) and political rights (71,72), the years of democratic

governance (73,74) and the presence of elections (75,76). While the vast majority of the studies

concluded that democracy is good for population health, some have not found any evidence

(70,71,77). Welfare state generosity, in terms of relatively high expenditures on health and social

services, is favorably associated with population health outcomes in cross-country studies in

developed (78–80) and developing countries (81,82). There are indications that (long-term) social

democratic (or left-wing) government participation has had a positive impact on some areas of

preventive health policy (83,84) and population health outcomes such as life expectancy and infant

mortality (53,85,86). Social democratic government has also been related to reduced health

inequalities (87) and spatial inequalities of healthcare services (88), but not all studies found a


15

positive relationship (89). Neoliberal policies, in contrast, were found to widen health inequalities

as a result of increased privatization and reduced welfare provisions (90). Moreover, in the

traditionalistic political culture in the Southern states of the United States, whose origin is found

in plantation-based economies, government is understood as a means to preserve the existing

social and economic order (91). To the present day, this political culture is associated with poorer

mortality outcomes compared to the moralistic political culture in Northern states, in which the

purpose of government is understood as a means to promote the general welfare of the population

(92). Globalization, defined by dependency indicators such as trade, foreign investment, and

national debt, has been negatively associated with population health (93), although the evidence of

this political determinants’ effect on health outcomes is least conclusive as studies are dispersed

across a diversity of outcomes (32).

Studies on political determinants of health that cover Caribbean states, in contrast, tend to focus

on the legacy of colonialism and the current division in political sovereignty status. Currently, 16

of 31 Caribbean states are politically sovereign, while the other states remain politically affiliated

to their former colonial administrators the Netherlands, France, the US and the UK. Colonialism

is defined as the policy or practice of acquiring political control over another country, occupying

it with settlers and exploiting it economically (57). In line with the negative undertone of this

definition, common consensus is that Western colonialism did more harm than good. There is

empirical evidence, however, that a longer duration of colonial experience is positively related to

both economic development and infant mortality (94). The contemporary version of “political

affiliation” also seems to make populations living in these territories not worse off: affiliated states

consistently outperform their sovereign counterparts in terms of health outcomes (95–97) and

economic development (94,96–99).

The previously mentioned evidence on the association of political affiliation with health and

economic development is mostly based on cross-sectional data from the 1980s onwards, so after

the main events of decolonization, the process in which a colony becomes politically independent.

Consequently, they do not shed light on when disparities between currently affiliated and

independent states came to be. To answer this question, Bertram used trade data to show that the

economic divergence between currently affiliated and sovereign states first became apparent

between the 1920’s and 1930’s and was well established prior to the first wave of decolonization

in the 1950’s (97). As the petition for independence in the Caribbean was primarily initiated by the

local governments of colonized countries and territories (100,101), this suggests that richer

Chapter 1

16

colonies have chosen to remain affiliated to their former colonizers, while poorer colonies have

opted for political sovereignty.

Figure 3 illustrates the almost clear-cut division in political sovereignty status for life expectancy

and economic development. For each Caribbean state, life expectancy at birth among affiliated

(grey triangle) and sovereign states (black square) is plotted against GDP per capita of that state.

The graph indicates that states that have remained affiliated to their Western colonizers generally

have better population health and higher economic development than states that have gained their

independence.

Given the concave relationship between life expectancy and GDP per capita, it is tempting to

conclude that political status had its effect on population health through economic development.

The association between the health and wealth of populations has long been established (102), and

remains important in the modern age (103). This, however, does not completely explain the

inequalities in life expectancies between currently sovereign and affiliated Caribbean states. Some

sovereign Caribbean states have been very successful in improving population health despite their

low economic development, such as high achiever Cuba. Others, with Trinidad and Tobago being

the most remarkable example, perform less well than might be expected based on their GDP per

Figure 3 Association between GDP per capita and life expectancy by political sovereignty status, 2015 Sources: UNdata, http://data.un.org and PAHO, PLISA database http://www.paho.org/data/index.php/en/


17

capita. This observation indicates that variations in determinants other than economic

development influence life expectancy development in Caribbean states as well. Whereas the

determinants of between-country health inequalities have been rather extensively studied in

Europe (see for example (89,104,105)), much less is known about the Caribbean region. The

variations in political status and population health in the Caribbean therefore offer interesting

opportunities for research on the political determinants of population health, so we can gain deeper

insights in why some governments have been more successful than others in improving the health

of their people.

This thesis

Shortcomings in the current literature

Despite increased recognition of the importance of political determinants for population health

improvements, the current literature has several shortcomings. First, there is a lack of comparative

health research among Western countries and their dependencies (also known as overseas

countries and territories, OCTs), the constituent countries of the Kingdom of the Netherlands in

particular. As we have previously mentioned, the health of Dutch Caribbean people living in the

Dutch Caribbean islands and in the Netherlands is poorer than that of the Dutch population. More

rigorous studies are needed to identify aspects that may inform health policy strategies to improve

the health of the Dutch Caribbean people.

Second, the number of studies on the political determinants of health within the Caribbean region

is scarce and typically involves cross-sectional comparisons. Given the fact that decision-making

on health occurs on the local level in -nearly- all Caribbean countries and territories, more insights

are needed in how political conditions shape population health in Caribbean states, and which

mediating pathways are involved. This needs to be better understood to identify factors that

hamper and stimulate population health improvements.

Lastly, a fundamental issue in all countries relates to how public funds should be used to invest in

population health, and which priorities should be set to achieve better health outcomes. This is

particularly important for Caribbean states, considering that their economies are fragile, their

human resources are limited and their health challenges require swift actions. Thus far, there are

hardly any studies on the impact of the implementation of “best practice” health interventions on

population health in the Caribbean region.

Chapter 1

18

Objectives and research questions

This thesis presents the findings of our efforts to understand why health outcomes in the Dutch

Caribbean islands are poorer than in the Netherlands and in several other Caribbean states and

proposes ways for reducing the between-country inequalities. The main objective of this thesis is

to provide a better insight into the health situation of the Dutch Caribbean, and the factors related

to this health situation, in particular the role of the political context and health policy performance.

This thesis addresses this aim specifically through the following research questions:

I. What is the health status of the Dutch Caribbean population?

II. To what extent do differences in population health in the Caribbean reflect differences in

political context and health policy performance?

In order to answer these research questions, we have made use of data that are derived from cross-

sectional health surveys, mortality registration systems, harmonized international databases, and

country reports. We use a quantitative observational approach and apply various statistical

methods to empirically examine patterns and trends of health inequalities between the Dutch

Caribbean, the Netherlands and other Caribbean territories. Moreover, we assess their association

with political conditions, policy implementation and national indicators to point the way to

effective public health policies. To address the multidisciplinary aspects of the determinants of

populations health, we cover theory from the fields of public health, medicine, political science,

organization science, economics, and sociology.

Outline

This thesis is divided in 7 chapters. Chapter 1 provides a general introduction into the topics

addressed in this thesis. It also describes the general aim, the research questions, and introduces

the methods used. After the general introduction and in line with the research questions, this thesis

is divided in two parts.

The first part consists of chapter 2 and 3, and focuses on the differences in population health

within the Kingdom of the Netherlands. In chapter 2, we compare self-reported health outcomes

and behaviors of Dutch Caribbean migrants living in the Netherlands to that of Dutch Caribbean

and Dutch non-migrants. The study in chapter 3 examines the contribution of deaths that are

amenable in the presence of timely and effective healthcare to the lower life expectancies in the

Dutch Caribbean in comparison to the Netherlands. The second part includes three chapters,


19

chapter 4, 5, and 6, and focuses on the political context of population health and health policy

performance in the Caribbean region. Specifically, in chapter 4 we investigate how life

expectancies in Caribbean states have evolved during the 1950-2014 period and evaluate whether

decolonization coincided with changes in life-expectancy, and similar changes in economic

development. In chapter 5, we assess whether differences in life expectancy trends between

Caribbean dependencies and their Western administrators are related to their degree of political

independence, and which causes of death contributed to life expectancy developments during the

1980-2014 period. Chapter 6 depicts a study in which we evaluate the health policy performance

of 16 Caribbean states in 11 different policy areas during the 2010-2015 period. We also explore

the association of the health policy performance score with national determinants and estimate the

potential health gains of “best-practice” health policies. We conclude this thesis with a general

discussion in chapter 7. Here we present our main findings, address methodical concerns,

interpret our results in the context of earlier studies, and propose ways for improving health

research and policy in the Dutch Caribbean islands.

Chapter 1

20

References

1. Verschuuren M, van Oers H, editors. Population Health Monitoring. Climbing the Information Pyramid. 1st ed. Cham: Springer International Publishing; 2019.

2. WHO Regional office for Europe. European Health Information Initiative [Internet]. WHO Europe; Available from: https://eupha.org/repository/advocacy/EHII_Booklet_EN_rev1.pdf

3. United Nations Department of Social and Economic Affairs. World Population Prospects: The 2017 Revision, DVD edition. [Internet]. 2017 [cited 2019 Apr 8]. Available from: http://esa.un.org/wpp/index.htm

4. World Health Organization. WHO mortality database, April 2018 update [Internet]. [cited 2018 Oct 1]. Available from: http://www.who.int/healthinfo/mortality_data/en/

5. Verstraeten S, Griffith M, Pin R. De Nationale Gezondheidsenquête 2017. Willemstad, Curaçao: Volksgezondheid Instituut Curaçao; 2018.

6. Dagevos J, Dagevos M. Minderheden meer gewicht. Over overgewicht bij Turken, Marokkanen, Surinamers en Antillianen en het belang van integratiefactoren. Den Haag: Sociaal en Cultureel Planbureau; 2008.

7. Lindeman E, Booi H, Cohen L. Monografie Antillianen 2010. De positie van Antillianen en Arubanen in Amsterdam. Amsterdam: Gemeente Amsterdam; 2010.

8. Schouten G, Stam B, Christiaanse B, Van Veelen J. Gezondheidsenquête 2008. Onderzoeksmethode en gezondheid in relatie tot achtergrondkenmerken. Rotterdam: GGD Rotterdam-Rijnmond; 2010.

9. Stirbu I, Kunst a E, Bos V, Mackenbach JP. Differences in avoidable mortality between migrants and the native Dutch in The Netherlands. BMC Public Health. 2006;6:78.

10. Gobièrnu di Kòrsou. Regeerprogramma 2017-2021. Ontplooien van Curaçao’s potentieel. Willemstad, Curaçao; 2017.

11. Gobièrno di Aruba. Regeerprogramma Kabinet Wever-Croes. Hunto pa Aruba. Oranjestad, Aruba; 2017.

12. Sint Maarten United Christian Democratic Coalition. Governing Program 2018-2022. Building a Sustainable Sint Maarten. Philipsburg, Sint Maarten; 2018.

13. Hoogers H. Tekst en Commentaar Grondwet, Commentaar op art. 43 Statuut Koninkrijk, Fundamentele rechten, rechtszekerheid en deugdelijkheid van bestuur. Deventer: Kluwer; 2015.

14. United Nations Department of Economic and Social Affairs Population Division. Changing levels and trends in mortality: the role of patterns of death by cause. ST/ESA/SER/A/318. New York: United Nations; 2012. 1–81 p.

15. Pan American Health Organization. Public Health Response to Chronic Diseases [Internet]. 26th Pan American Sanitary Conference, 54th session of the Regional Committee, CSP26/15. Washington, USA; 2002. Available from: http://iris.paho.org/xmlui/handle/123456789/21540

16. Europe PMC Funders Group Global. Global, regional and national prevalence of overweight and obesity in children and adults 1980-2013: A systematic analysis. Lancet. 2014;384(9945):766–81.

17. Pan American Health Organization (PAHO/WHO), Caribbean Community Secretariat (CARICOM). Report of the Caribbean commission on health and development. Kingston, Jamaica: Ian Randle Publishers; 2006.


21

18. Healthy Caribbean Coalition. Responses to NCDs in the Caribbean community [Internet]. 2014. Available from: https://ncdalliance.org/sites/default/files/resource_files/HCC NCDA RSR FINAL.pdf

19. United Nations Office on Drugs, Latin America and Caribbean region of the World Bank. Report No. 37820. Crime, violence, and development: trends, costs, and policy options in the Caribbean. United Nations and World Bank; 2007.

20. Institute for Health Metrics and Evaluation. Global Burden of Disease [Internet]. 2019 [cited 2019 Aug 4]. Available from: https://vizhub.healthdata.org/gbd-compare/

21. Everest-Phillips M. Small, so simple? Complexity in Small Island Developing States. UNDP Global Centre for Public Service Excellence. Singapore: United Nations; 2014.

22. Bliss KE. Health in Latin America and the Caribbean. Challenges and opportunities for US engagement. Washington: CSIS Global Health Policy Center; 2009.

23. Pan American Health Organization. PLISA Health Information Platform for the Americas [Internet]. 2019 [cited 2019 Aug 4]. Available from: http://www.paho.org/data/index.php/en/indicators/visualization.html

24. Cloos P. Health inequalities in the Caribbean: increasing opportunities and resources. Glob Health Promot. 2010;17(1):73–6.

25. McKeown T. The role of medicine. Dream, mirage or nemesis? London: The Nuffield Provincial Hospitals Trust; 1976.

26. McGinnis JM, Williams-Russo P, Knickman JR. The case for more active policy attention to health promotion. Health Aff. 2002;21(2):78–93.

27. McKeown T, Record RG, Turner RD. An interpretation of the decline of mortality in England and Wales during the twentieth century. Popul Stud (NY). 1975;29(3):391–422.

28. Pan American Health Organization (PAHO/WHO). Strategic Plan 2014-2019. Championing Health: Sustainable Development and Equity. 345th ed. PAHO/WHO; 2014.

29. Virchow R. Der Armenarzt. Die Med Reform. 1848;18(S125):125–7.

30. Engels F. The condition of the working-class in England. Germany: Otto Wigand; 1845.

31. Chadwick E. Report on the sanitary condition of the labouring population of Great Britain. London, United Kingdom: House of Commons; 1842.

32. Barnish M, Tørnes M, Nelson-Horne B. How much evidence is there that political factors are related to population health outcomes? An internationally comparative systematic review. BMJ Open. 2018;8(e020886).

33. Marmot M. Social determinants of health inequalities. Lancet. 2005;365(9464):1099–104.

34. World Health Assembly 66. Social determinants of health: report by the secretariat. [Internet]. World Health Organization. WHO; 2013. Available from: https://apps.who.int/iris/handle/10665/149438

35. Braveman P, Gottlieb L. The Social Determinants of Health: It’s Time to Consider the Causes of the Causes. Public Health Rep. 2017;129(1_suppl2):19–31.

36. Wilkinson RG, Marmot MG, editors. Social determinants of health: the solid facts. [Internet]. World Health Organization Europe; 2003. Available from: http://www.euro.who.int/__data/assets/pdf_file/0005/98438/e81384.pdf

Chapter 1

22

37. Graham H. Social determinants and their unequal distribution: clarifying policy understandings. Milbank Q. 2004;82(1):101–24.

38. Smith JA, Zhao W, Wang X, Ratliff SM, Mukherjee B, Kardia SLR, et al. Neighborhood characteristics influence DNA methylation of genes involved in stress response and inflammation: the multi-ethnic study of atherosclerosis. Epigenetics. 2017;12(8):662–73.

39. Mackenbach JP. Socioeconomic inequalities in health in high-income countries: the facts and the options. In: Oxford Textbook of Global Public Health. 2015.

40. Bambra C, Fox D, Scott-Samuel A. Towards a politics of health. Health Promot Int. 2005;20(2):187–93.

41. Carpenter D. Is Health Politics Different? Annu Rev Polit Sci. 2012;15(1):287–311.

42. Greer SL, Bekker M, De Leeuw E, Wismar M, Helderman JK, Ribeiro S, et al. Policy, politics and public health. Eur J Public Health. 2017;27(ii):40–3.

43. Ståhl T, Wismar M, Ollila E, Lahtinen E, Leppo K. Health in All Policies. Prospects and potentials. Helsinki: European Observatory on Health Systems and Policies; 2006.

44. Shah PS, Balkhair T, Ohlsson A, Beyene J, Scott F, Frick C. Intention to become pregnant and low birth weight and preterm birth: A systematic review. Matern Child Health J. 2011;15(2):205–16.

45. Blecher E. The impact of tobacco advertising bans on consumption in developing countries. J Health Econ. 2008;27(4):930–42.

46. Anderson P, Chisholm D, Fuhr DC. Effectiveness and cost-effectiveness of policies and programmes to reduce the harm caused by alcohol. Lancet. 2009;373(9682):2234–46.

47. Margie P, Scurfield R, Sleet D, Mohan D, Hyder AA, Jarawan E, et al., editors. World report on road traffic injury prevention. Geneva: World Health Organization; 2004.

48. Redondo M, Hernández-Aguado I, Lumbreras B. The impact of the tax on sweetened beverages: A systematic review. Am J Clin Nutr. 2018;108(3):548–63.

49. Lauby-Secretan B, Scoccianti C, Loomis D, Benbrahim-Tallaa L, Bouvard V, Bianchini F, et al. Breast-cancer screening: viewpoint of the IARC working group. N Engl J Med. 2015;372(24):2353–8.

50. International Agency for Research on Cancer. Cervix cancer screening. IARC Handbook of Cancer Prevention volume 10. Lyon; 2005.

51. Plug I, Hoffmann R, Mackenbach JP, editors. Avoidable mortality in the European Union. Volume 1: Final report. Public Health. Rotterdam: EU Public Health Program 2007106; 2011.

52. Navarro V, Muntaner C, Borrell C, Benach J, Quiroga A, Rodríguez-Sanz M, et al. Politics and health outcomes. Lancet. 2006 Sep 16;368(9540):1033–7.

53. Navarro V, Shi L. The political context of social inequalities and health. Soc Sci Med. 2001;52(3):481–91.

54. Pickett KE, Wilkinson RG. Child wellbeing and income inequality in rich societies: ecological cross sectional study. BMJ. 2007;335(7629):1080.

55. Pickett KE, Wilkinson RG. Income inequality and health: a causal review. Soc Sci Med. 2015;128:316–26.

56. Oliver TR. The politics of public health policy. Annu Rev Public Health. 2006;27(1):195–233.


23

57. McLean I, McMillan A. The Concise Oxford Dictionary of Politics. 4th ed. Oxford University Press; 2013.

58. Wikipedia. Politics [Internet]. 2019 [cited 2019 May 1]. Available from: https://simple.wikipedia.org/wiki/Politics

59. Franco A, Alvarez-Dardet C, Ruiz MT. Effect of democracy on health: ecological study. BMJ. 2004;329:1421–3.

60. Besley T, Kudamatsu M, Merlo A, Olken B, Nunn N. Health and democracy. Am Econ Rev. 2006;96(2):313–8.

61. Bollyky TJ, Templin T, Cohen M, Schoder D, Dieleman JL, Wigley S. The relationships between democratic experience, adult health, and cause-specific mortality in 170 countries between 1980 and 2016: an observational analysis. Lancet. 2019;393(10181):1628–40.

62. Dolea C, Nolte E, McKee M. Changing life expectancy in Romania after the transition. J Epidemiol Community Health. 2002 Jun;56(6):444–9.

63. Parmar PK, Barina CC, Low S, Tun KT, Otterness C, Mhote PP, et al. Health and human rights in eastern Myanmar after the political transition: a population-based assessment using multistaged household cluster sampling. PLoS One. 2015 Jan;10(5):e0121212.

64. Varnik A, Kolves K, Sisask M, Samm A, Wasserman D. Suicide Mortality And Political Transition: Russians In Estonia Compared To The Estonians In Estonia And To The Population Of Russia. Trames. 2006;10(3):268–77.

65. Leon D, Chenet L, Shkolnikov V, Zakharov S, Shapiro J, Rakhmanova G, et al. Huge variation in Russian mortality rates 1984-94: artefact, alcohol, or what? Lancet. 1997;350(9075):383–8.

66. Brainerd E. Market Reform and Mortality in Transition Economies. World De. 1998;26(11):2013–27.

67. Nolte E, Shkolnikov V, McKee M. Changing mortality patterns in East and West Germany and Poland. II: short-term trends during transition and in the 1990s. J Epidemiol Community Health. 2000;54(12):899–906.

68. Mackenbach JP, Hu Y, Looman CW. Democratization and life expectancy in Europe, 1960-2008. Soc Sci Med. 2013;93:166–75.

69. Lake DA, Baum MA. The invisible hand of democracy: political control and the provision of public services. Comp Polit Stud. 2001;34(6):587–621.

70. Gauri V, Khaleghian P. Immunization in developing countries: its political and organizational determinants. World Dev. 2002;30(12):2109–32.

71. Houweling TAJ, Caspar AEK, Looman WN, Mackenbach JP. Determinants of under-5 mortality among the poor and the rich: A cross-national analysis of 43 developing countries. Int J Epidemiol. 2005;34(6):1257–65.

72. Stroup MD. Economic Freedom, Democracy, and the Quality of Life. World Dev. 2007;35(1):52–66.

73. Ross M. Is democracy good for the poor? Am J Pol Sci. 2006;50(4):860–74.

74. Gerring J, Thacker SC, Alfaro R. Democracy and Human Development. J Polit. 2012;74(1):1–17.

75. Zweifel TD, Navia P. Democracy, Dictatorship, and Infant Mortality. J Democr. 2000;11(2):99–114.

Chapter 1

24

76. Navia P, Zweifel TD. Democracy, Mortality, and Infant Mortality Revisited. J Democr. 2003;14(3):90–103.

77. Michael R. Is Democracy Good for the Is Democracy Good for the Environment ? Am J Pol Sci. 2006;50(4):860–74.

78. Bradley EH, Elkins BR, Herrin J, Elbel B. Health and social services expenditures: Associations with health outcomes. BMJ Qual Saf. 2011;20(10):826–31.

79. Dahl E, van der Wel KA. Educational inequalities in health in European welfare states: a social expenditure approach. Soc Sci Med. 2013;81:60–9.

80. Bremberg SG. The rate of country-level improvements of the infant mortality rate is mainly determined by previous history. Eur J Public Health. 2016;26(4):597–601.

81. Novignon J, Olakojo SA, Nonvignon J. The effects of public and private health care expenditure on health status in sub-Saharan Africa: new evidence from panel data analysis. Health Econ Rev. 2012;2(1):22.

82. Akinci F, Hamidi S, Suvankulov F, Akhmedjonov A. Examining the impact of health care expenditures on health outcomes in the Middle East and North Africa (MENA) region. J Heal Care Financ. 2014;41(1):1–23.

83. Mackenbach JP, McKee M. Social-democratic government and health policy in Europe: a quantitative analysis. Int J Heal Serv. 2013;43(3):389–413.

84. Bosdriesz JR, Willemsen MC, Stronks K, Kunst AE. Tobacco control policy development in the European Union: do political factors matter? Eur J Public Health. 2015;25(2):190–4.

85. Chung H, Muntaner C. Political and welfare state determinants of infant and child health indicators: An analysis of wealthy countries. Soc Sci Med. 2006;63(3):829–42.

86. Lin R-T, Chen Y-M, Chien L-C, Chan C-C. Political and social determinants of life expectancy in less developed countries: a longitudinal study. BMC Public Health. 2012;12(1):85.

87. Navarro V, Muntaner C, Borrell C, Benach J, Quiroga Á, Rodríguez-Sanz M, et al. Politics and health outcomes. Lancet. 2006;368(9540):1033–7.

88. Bennema-Broos M, Groenewegen PP, Westert GP. Social democratic government and spatial distribution of health care facilities: the case of hospital beds in Germany. Eur J Public Health. 2001;11(2):160–5.

89. Mackenbach JP, McKee M. A comparative analysis of health policy performance in 43 European countries. Eur J Public Health. 2013;23(2):195–201.

90. Scott-Samuel A, Bambra C, Collins C, Hunter DJ, McCartney G, Smith K. The impact of Thatcherism on health and well-being in Britain. Int J Heal Serv. 2014;44(1):53–71.

91. Elazar DJ. American Federalism: A View from the States. Second edi. New York: Crowell Company; 1972.

92. Kunitz SJ. The health of populations: general theories and particular realities. First edit. Oxford University Press; 2006.

93. Muntaner C, Borrell C, Ng E, Chung H, Espelt A, Rodriguez-Sanz M, et al. Politics, welfare regimes, and population health: controversies and evidence. Sociol Health Illn. 2011;33(6):946–64.

94. Feyrer J, Sacerdote B. Colonialism and Modern Income: Islands as Natural Experiments. Rev Econ Stat. 2009;91(2):245–62.


25

95. Mcelroy JL, Sanborn K. The Propensity for Dependence in Small Caribbean and Pacific Islands. Bank Valletta Rev. 2005;(31):1–16.

96. McElroy JL, Pearce KB. The advantages of political affiliation: Dependent and independent small-island profiles. Round Table Commonw J Int Aff. 2006;95(386):529–39.

97. Bertram G. Economic consequences of decolonisation in small island economies: a long-run analysis. In Reunion Island: Proceedings of the conference "Tourism specialization and vulnerability: evidence and challenges for sustainable development in small island territories”; 2014.

98. Armstrong HW, Read R. The economic performance of small states and islands: the importance of geography. In Taiwan: Proceedings of the eighth Islands of the World International Conference; 2004. p. 721–48.

99. Armstrong HW, Read R. Comparing the economic performance of dependent territories and sovereign microstates. Econ Dev Cult Change. 2000;48(2):285–306.

100. Oostindie G, Klinkers I. Decolonising the Caribbean. Dutch policies in a comparative perspective. Amsterdam: Amsterdam University Press; 2003.

101. Elkins Z, Ginsburg T, Blount J. Constitutional reform in the English-speaking Caribbean: challenges and prospects. Proceedings in the “Conflict Prevention and Peace Forum”; 2011.

102. Preston SH. The changing relation between mortality and level of economic development. Popul Stud A J Demogr. 1975;29(2):231–48.

103. Mackenbach JP, Looman CW. Life expectancy and national income in Europe, 1900-2008: an update of Preston’s analysis. Int J Epidemiol. 2013;42(4):1100–10.

104. Thomson K, Bambra C, Mcnamara C, Huijts T, Todd A. The effects of public health policies on population health and health inequalities in European welfare states: protocol for an umbrella review. Syst Rev. 2016;5(57).

105. UCL Institute of Health Equity. Review of social determinants and the health divide in the WHO European region: final report. Copenhagen: WHO Regional Office for Europe; 2014.

Part I Population health comparisons

within the Kingdom of the

Netherlands

Chapter 2

The health of Antillean migrants in the Netherlands:

a comparison with the health of non-migrants in both the

countries of origin and

destination

SPA Verstraeten, CL van den Brink, JP Mackenbach and HAM van Oers

International Health 2018, vol. 10, Iss. 4, pp. 258-267

https://doi: 10.1093/inthealth/ihy026

Chapter 2

30

Abstract

Background We examined risk factor and health differences between Antillean migrants in

the Netherlands and Antillean and Dutch non-migrants, and related our findings to four

commonly used explanations for migrant health disparities.

Methods We used nationally representative data from the 2012 Dutch Public Health Monitor

and the 2013 National Health Survey Curaçao. We calculated weighted rates and assessed

significance using chi-square. We used logistic regression analyses to compare health behaviors

and outcomes between Antillean migrants and the non-migrant populations.

Results Overall, Antillean migrants had poorer physical and mental health than Antillean and

Dutch non-migrants. For overweight/obesity and tobacco and alcohol use, Antillean migrants

had rates in-between those of the Antillean and Dutch non-migrants. The poor health of

Antillean migrants persisted in the second-generation who were born in the Netherlands.

Conclusions Patterns of differences in physical and mental health among the study

populations were suggestive of a ‘stressful environment’ effect. The poorer health of Antillean

migrants may be partly determined by host-country specific stressors such as perceived

discrimination, spatial concentration in multi-ethnic neighborhoods and reduced social mobility.

The health of Antillean migrants in the Netherlands

31

Introduction

Migrant health studies in developed countries have primarily focused on the ‘host country’

perspective of the country of destination and commonly describe in what ways migrants’ health

differs from the non-migrant population. The literature offers a number of explanations for health

disparities between migrants and non-migrants, and a common distinction is made between

proposals of theories that take the temporal dimension of migrant health into account (e.g. health

transition theory (1), cumulative disadvantage theory (2), and the life course approach (3), and

hypotheses that are applied to explain cross-sectional findings.

Many cross-sectional health studies focus on the paradoxical finding that some migrant groups

have better health outcomes than their respective non-migrant population, at least initially upon

arrival in the host country. The ‘healthy migrant’ effect (or the related ‘Hispanic Paradox’) proposes

that migration is selective for younger and healthier individuals (4,5), for example through positive

self-selection of healthy individuals or preferential immigrant health policies of the host country

(6,7). For most migrant groups, however, chronic health problems are more common than among

the non-migrant host populations in developed countries (8–10). Several theories have been

proposed to explain these findings. The ‘affluent environment’ effect suggests that migrants’

increased risks for obesity and diabetes is caused by relocation to a more affluent environment and

the adoption of sedentary lifestyles with calorie-dense diets (11,12). In contrast, the ‘stressful

environment’ effect describes the high incidence of psychiatric disorders and proposes that

migrants experience high levels of (psycho)social stress caused by their minority status and socio-

economic disadvantage (13,14). The explanation of ‘convergence’ takes into account that migrant

health disparities are not static, but change over time (15,16). This hypothesis implies that migrant’s

health profiles will eventually, based on length of stay or successive generations, converge towards

the host country’s population norm. The direction of convergence can range from more favorable

initial levels to the less favorable levels in the host country (upwards convergence), or from less

favorable initial levels to the more favorable levels in the host country (downwards convergence)

(9).

The commonly proposed theories have advanced our understanding of migrant health disparities,

but their actual significance to inform policy development is unclear due to fragmentary and

sometimes inconsistent evidence (9). Moreover, a weakness in the migrant health literature is that

many studies do not take into account the ways a migrant group’s health status differs from the

population in the country of origin. Without this ‘home country’ perspective on migrant health, it

is not possible to differentiate whether the disparities between a migrant group and a respective

Chapter 2

32

non-migrant population are the result of biologically or culturally determined differences, or are in

fact the result of host country specific circumstances that disproportionally affect migrants more

than non-migrant populations. The purpose of our current study was therefore to address this

paucity in the migrant health literature and to compare the health of Antillean migrants in the

Netherlands with non-migrants in both the countries of origin and destination.

Antillean migrants in the Netherlands represent a unique migrant population. Antilleans from

Curaçao and the five other, less populous, Dutch Caribbean islands acquire Dutch nationality at

birth, which allows them to freely settle and receive the same social and educational benefits as the

non-migrant Dutch in the Netherlands. In fact, the Dutch Caribbean islands from which they

originate have been part of the Kingdom of the Netherlands since the 17th century and still have

strong (socio-)economic, political and legal ties to its former colonizer today. The Dutch Caribbean

islands use Papiamento, Dutch and English as official languages and their systems of education

and healthcare are largely based on those in the Netherlands, albeit of undetermined quality.

Nonetheless, Antillean migrants are considered a non-western migrant minority population in the

Netherlands, both in political rhetoric and by registration in the Municipal Personal Records

Database (GBA). Antilleans are the fourth largest non-Western migrant population in the

Netherlands, and although their numbers represent only a small proportion (0.9%) of the Dutch

population, they represent 50% of the Antilleans on the Dutch Caribbean islands.

The purpose of our current study was to compare the health status and behaviors of Antillean

migrants in the Netherlands with Antillean non-migrants in Curaçao and Dutch non-migrants in

the Netherlands and to determine whether health disparities persist in second-generation Antillean

migrants who were born in the Netherlands. Then, we relate our findings to the common

explanations of migrant health disparities in order to assess which one provides the most plausible

explanation for the health outcomes of Antillean migrants relative to Dutch non-migrants.

Methods

Data sources

We used data from two population-based surveys that collected data on background factors, health

behaviors and self-reported health among the non-institutionalized population in the Netherlands

and in Curaçao. Dutch data were from the 2012 Public Health Monitor and Curaçao data were

from the 2013 National Health Survey. We selected comparable questions, cross-standardized

categories according to DPHM2012 categories and combined the datasets to ensure comparability.


33

Both surveys included sample weights to assure that estimates are representative at the national

level. Detailed descriptions of the health surveys were published elsewhere and are briefly

described below (17,18).

The Dutch Public Health Monitor (DPHM) 2012

The Dutch Public Health Monitor’s objective is to provide a comprehensive overview of health

status, social and lifestyle factors of the adult population in the Netherlands for regional and

national information needs. In 2012, data were collected from 378.195 respondents and included

1.556 Antillean respondents. Migrant background was indicated according to the Municipal

Personal Records Database (GBA). The majority of data (97.2%) were collected by 28 Community

Health Service departments (GGD’s) using digital or written questionnaires during the fall of 2012.

Remaining data were collected by Statistics Netherlands using digital or written questionnaires or

face-to-face interviews. Sample weights correct for gender, age, marital status, urbanization,

household size, ethnicity, income, municipality and GGD-region.

National Health Survey Curaçao (NHSC) 2013

The NHSC’s objective is to collect nationally representative data on the health status, health

determinants and healthcare use of the adult population in Curaçao. The questionnaire was based

on international constructs to ensure validity and improve international comparability. Data

collection was organized in collaboration with Statistics Curaçao. Data were collected from 3000

respondents using face-to-face interviews conducted in Papiamento, Dutch, English or Spanish

during January and February 2013. Sample weights correct for gender and age.

Study populations

Antillean non-migrants participated in the NHSC2013 and were born in Curaçao or one of the

other Dutch Caribbean islands. Antillean migrants and Dutch non-migrants participated in the

DPHM2012. First-generation Antillean migrants were born on the Dutch Caribbean islands.

About three-quarters were born and -partly- raised in Curaçao (19). Second-generation Antillean

migrants were born in the Netherlands and had at least one Antillean parent. Dutch non-migrants

were born in the Netherlands and had parents who were born in the Netherlands as well.

Risk factor and health status variables

Our analysis focused on comparable variables between the surveys. For risk factors, this included

overweight (BMI 25), obesity (BMI 30), two indicators for tobacco use -smoker and heavy

Chapter 2

34

smoker (>20 cigarettes/day)-, three indicators for alcohol use -alcohol user, excessive user (>21

(men) or >14 (women) consumptions per week) and heavy user (>5 (men) or >3 (women)

consumptions on one day, at least weekly)- and one variable for physical inactivity (0 days during

the past 7 days comparable for DPHM2012 respondents only).

For health status, variables included self-reported (very) bad health, morbidity (captured by

respondents’ reports of stroke, heart infarction, other severe heart disease, cancer, diabetes,

migraine and hypertension, with diabetes, migraine and hypertension assessed separately as well),

multi-morbidity (two or more of the previously mentioned diseases), and limitations in sight,

hearing or mobility).

Psychological distress was assessed with questions from the RAND 36-Item Short Form

(NHSC2013) and the Kessler Psychological Distress Scale (DPHM2012). Based on five questions

that were comparable across these instruments, we calculated the variable psychological

complaints according to the Mental Health Inventory method (20).

Covariates

We adjusted for individual-level sociodemographic characteristics that are known determinants of

health during the logistic regression analysis. Demographic characteristics included a continuous

variable for age and binary variable for marital status (living together, unmarried, divorced or

widowed, with married as reference group). Available for DPHM2012 respondents only, this also

included an urbanization variable (moderately urban, not urban, with (very) strongly urban as

reference group). This variable is based on the address density of the surrounding area as defined

by Statistics Netherlands and retrospectively added to the dataset based on the postal code of the

respondent. Socio-economic characteristics included binary variables for educational status (low

(ISCED 0-1), intermediate1 (ISCED 2-3), or 2 (ISCED 4), with high education (ISCED 5 and 6)

as reference group) and employment (paid work, yes/no).

Statistical analysis

For each gender separately, we computed descriptive statistics with use of the respective

DPHM2012 or NHSC2013 sample weights. Significance of differences for categorical variables

was tested with chi-square. We estimated odds ratios between study populations with logistic

regression, using Antillean migrants as the reference group, with and without adjustment for

sociodemographics. Next, intergenerational differences among Antillean migrants were assessed.

Considering that the second-generation Antillean migrant group was small and younger, we


35

combined both genders and included 19- to 65-year olds only to estimate differences between first

and second-generation migrants, using the same regression models with the addition of gender.

95% confidence intervals were used to assess significance. Data were analyzed using SPSS-version

20.

Interpretation of the logistic regression results

In order to relate our findings to common explanations of migrant health disparities, we compared

patterns of differences between Antillean migrants and non-migrant populations with the

differences that one would expect if each of these explanations applied. Although the explanations

are not mutually exclusive, the patterns of differences among the study-populations indicate

whether certain explanations hold true across multiple health outcomes.

Table 1 summarizes our expectations under each of the explanations. If health problems are less

common among Antillean migrants compared to non-migrant populations, the ‘healthy migrant’

effect would be a plausible explanation. Conversely, since Curaçao has fewer economic resources

than the Netherlands, higher prevalence of overweight/obesity and diabetes among Antillean

migrants is in accordance with the ‘affluent environment’ effect. Similarly, higher rates for mental

and subjective health related variables are suggestive for the ‘stressful environment’ effect. We

expect that rates of Antillean migrants are in-between rates of the non-migrant populations in the

case of ‘convergence’. Additionally, Antilleans may have biologically or culturally determined risk,

or protective, factors that persist after migration. If so, we expect no significant differences among

the Antillean populations, but significant disparities relative to the Dutch non-migrants.

Chapter 2

36

Table 1 Interpretation of odds ratio’s (OR) from logistic regression analysis Results logistic regression analysis Interpretation Antillean non-migrants in Curaçao

Antillean migrants in the Netherlands

Dutch non-migrants in the Netherlands Finding in accordance with:

OR more favorable 1.0 OR more favorable Stressful or affluent environmenta OR more favorable 1.0 OR less favorable Upwards convergence OR more favorable 1.0 OR similar Upwards convergenceb

OR similar 1.0 OR more favorable Biological or cultural risk factors of Antilleans that exist after migration

OR similar 1.0 OR less favorable Biological or cultural protective factors of Antilleans that exist after migration

OR similar 1.0 OR similar No effect OR less favorable 1.0 OR more favorable Downwards convergence OR less favorable 1.0 OR less favorable Healthy migrant effect OR less favorable 1.0 OR similar Downwards convergence (?)b,c a Depending on health outcome, it may be argued that either the affluent or stressful environment explanation applies. b not directly in accordance with given explanation because rates of Antillean migrants in the Netherlands may not reflect the end-point of convergence, but this could be reasoned with additional arguments. c May also be in accordance with the healthy migrant effect.

Results

Table 2 presents sociodemographic characteristics of the study populations. For both genders, we

found significant differences between Antillean migrants, Antillean non-migrants and Dutch non-

migrants for all characteristics. Compared with Antillean non-migrants, Antillean migrants were

younger, higher educated, and more likely to be cohabitating and perform paid work. Antillean

migrants were also younger than Dutch non-migrants, but less likely to be married, employed, and

more likely to live in an urban environment. Antillean migrant women were higher educated than

Dutch non-migrants women, whereas Antillean migrant men were lower educated than Dutch

non-migrants men. The health characteristics of the study populations are presented in table A1

of the Appendix.

The logistic regression results (figure 1 and Appendix table A2 and A3) showed that health status

and risk factors varied substantially between Antillean migrants, Antillean non-migrants and Dutch

non-migrants. In the models adjusted for age, marital status, educational level and employment,

overall health status of Antillean migrants was unfavorable compared with both non-migrant

populations (figure 1). Addition of the urbanization variable did not affect the significance of

differences between Antillean migrants and their Dutch counterparts (Appendix table A2 and A3).


37

Table 2 Sociodemographic characteristics of the given populations, by gender Men

Antilleans non-migrants (n=907)

Antilleans migrants (n=649)

Dutch non-migrants (n=153.006)

Age in years (range) 48.8 (19-93) 40.1 (19-93) 49.1 (19-107) Age in categories in n (weighted %)a,b

19-24 years 53 (11.4) 86 (17.9) 7.449 (9.7) 25-44 years 148 (28.7) 228 (46.6) 27.850 (32.4) 45-65 years 381 (39.8) 195 (27.8) 48.942 (37.1) 65 years and older 325 (20.2) 140 (7.6) 68.765 (20.8) Education level in n (weighted %)a,b Low 207 (18.3) 54 (6.6) 12.490 (5.8) Intermediate 1 362 (43.5) 209 (31.2) 46.136 (26.2) Intermediate 2 273 (31.5) 202 (34.9) 45.081 (34.7) High 65 (6.7) 160 (27.3) 44.804 (33.3) Missing 0 24 4.495 Marital status in n (weighted %)a,b Married/registered partnership 466 (43.9) 220 (29.1) 101.988 (57.3) Living together 66 (9.1) 83 (19.4) 11.576 (13.6) Unmarried 258 (38.0) 211 (41.8) 17.526 (21.0) Divorced 67 (5.7) 79 (7.4) 7.137 (5.0) Widow(er) 50 (3.3) 16 (2.3) 8.286 (3.1) Missing 0 40 6.493 Urbanization in n (weighted %)b (Very) strongly urban - 518 (84.0) 54.307 (42.7) Moderately urban - 62 (9.1) 29.472 (21.0) (very) Little urban - 69 (6.9) 69.227 (36.3) Employment in n (weighted %)a,b Yes 421 (58.4) 352 (65.7) 73.237 (69.2) No 486 (41.6) 257 (34.3) 71.432 (30.8) Missing 0 24 8.337

Women Antillean non-migrants (n = 1.549)

Antilleans migrants (n = 907)

Dutch non=migrants (n = 182.096)

Age in years (range) 50.1 (19-96) 41.8 (19-92) 50.7 (19-103) Age in categories in n (weighted %)a,b

19-24 years 61 (10.0) 151 (15.5) 10.486 (9.1) 25-44 years 327 (28.3) 286 (43.4) 37.629 (30.4) 45-65 years 646 (39.7) 270 (32.3) 56.407 (35.6) 65 years and older 515 (22.0) 200 (8.8) 77.574 (24.9) Education level in n (weighted %)a,b Low 437 (22.3) 68 (5.5) 17.350 (7.0) Intermediate 1 587 (39.4) 287 (26.4) 70.446 (32.9) Intermediate 2 414 (30.7) 285 (39.2) 46.957 (31.3) High 111 (7.6) 222 (28.8) 41.336 (28.7) Missing 0 45 6.007 Marital status in n (weighted %)a,b Married/registered partnership 439 (26.9) 235 (26.9) 101.793 (54.1) Living together 90 (6.5) 93 (16.9) 15.436 (13.3) Unmarried 598 (45.2) 350 (40.6) 17.912 (15.2) Divorced 202 (11.2) 133 (13.2) 11.490 (7.1) Widow(er) 220 (10.2) 37 (2.3) 27.901 (10.3) Missing 59 7.564 Urbanization in n (weighted %)b (Very) strongly urban - 702 (80.4) 66.375 (43.0) Moderately urban - 108 (12.8) 34.755 (21.2) (very) Little urban - 97 (6.8) 80.966 (35.8) Employment in n (weighted %)a,b Yes 626 (47.0) 420 (58.1) 77.756 (59.2) No 923 (53.0) 426 (41.9) 89.628 (40.8) Missing 0 45 14.712 a Antillean non-migrants are significantly different from Antillean migrants based on the Chi-square test (p = <0.001) b Dutch non-migrants are significantly different from Antillean migrants based on the Chi-square test (p = <0.001)

Chapter 2

38

The poorer health of Antillean migrants relative to the non-migrant populations did not resemble

generally worse health behaviors of this group. Overweight/obesity was less (for women) or

similarly (for men) common among Antillean migrants than among Antillean non-migrants, but

more common than among Dutch non-migrants (figure 1). Antillean migrants’ rates for tobacco

and alcohol use were in-between the lower rates of Antillean non-migrants and the higher rates of

Dutch non-migrants. Physical inactivity was more common among Antillean migrants than among

Dutch non-migrants.

Figure 2 shows the results of intergenerational differences among Antillean migrants (see

Appendix table A4 for full results). While limitations and hypertension were less prevalent among

second than among first-generation Antillean migrants in the adjusted models, other health

problems did not differ significantly. Risk factor prevalence among second-generation Antillean

migrants converged further to the Dutch population norm.

Discussion

Summary of main findings

Overall, Antillean migrants had poorer physical and mental health status than Antillean and Dutch

non-migrants. For overweight/obesity and tobacco and alcohol use Antillean migrants had rates

in-between those of Antillean and Dutch non-migrants. The poor health of Antillean migrants

persisted in the second-generation who were born in the Netherlands.

Strengths and limitations

Compared with similar studies on migrant health disparities, strengths of our study are that 1. We

study Antillean migrants who are Dutch nationals by birth, which allows them to freely settle in

the Netherlands and entitles them to the same social and educational benefits non-migrant Dutch

receive in the Netherlands, 2. the health of migrants was compared relative to that of non-migrants

in both country of origin and destination, and 3. a broad variety of health variables was examined.

The

hea

lth o

f Ant

illea

n m

igra

nts

in t

he N

ethe

rlan

ds

39

Fig

ure

1D

ispar

ity b

etw

een

Ant

illea

n m

igra

nts i

n th

e N

ethe

rland

s and

the

spec

ified

stud

y po

pulat

ion

by g

ende

r. M

odel

adju

sted

for

age,

mar

ital s

tatu

s, ed

ucat

iona

l lev

el an

d em

ploy

men

t. O

R=O

dds

ratio

, BM

I=Bo

dy M

ass

Inde

x, 1

>21

(m

en)

or >

14 (

wom

en)

cons

umpt

ions

/wee

k, 2 >

5 (m

en) o

r >3

(wom

en) c

onsu

mpt

ions

on

one

day/

wee

k, 3

BMI

25),

4 BM

I30

, 5 C

aptu

red

by st

roke

, hea

rt in

farc

tion,

oth

er s

ever

e he

art

dise

ase,

canc

er, d

iabet

es, m

igra

ine

and/

or h

yper

tens

ion,

one

or

mor

e, 6

Capt

ured

by

stro

ke, h

eart

infa

rctio

n, o

ther

seve

re h

eart

dise

ase,

canc

er, d

iabet

es, m

igra

ine

and/

or h

yper

tens

ion,

two

or m

ore.

Cha

pter

2

40

Fig

ure

1(c

ontin

ued)

The

hea

lth o

f Ant

illea

n m

igra

nts

in t

he N

ethe

rlan

ds

41

Fig

ure

2 E

stim

ated

ass

ociat

ion

betw

een

first

-gen

erat

ion

Ant

illea

n m

igra

nts i

n th

e N

ethe

rland

s and

the

spec

ified

stud

y po

pulat

ion

by g

ende

r, 19

- to

65-y

ear o

lds.

Mod

el ad

just

ed fo

r age

, mar

ital s

tatu

s, ed

ucat

iona

l lev

el an

d em

ploy

men

t. O

R=O

dds r

atio

, BM

I=Bo

dy M

ass I

ndex

, 1>

21

(men

) or >

14 (w

omen

) con

sum

ptio

ns/w

eek,

2 >5

(men

) or >

3 (w

omen

) con

sum

ptio

ns o

n on

e da

y/w

eek,

3 BM

I25

), 4 BM

I30

, 5 C

aptu

red

by

stro

ke, h

eart

infa

rctio

n, o

ther

sev

ere

hear

t dise

ase,

canc

er, d

iabet

es, m

igra

ine

and/

or h

yper

tens

ion,

one

or m

ore,

6 Ca

ptur

ed b

y st

roke

, hea

rt in

farc

tion,

oth

er se

vere

hea

rt di

seas

e, ca

ncer

, diab

etes

, mig

rain

e an

d/or

hyp

erte

nsio

n, tw

o or

mor

e.

Cha

pter

2

42

Fig

ure

2 (c

ontin

ued)


43

Our study’s main limitation is that the NHSC2013 (Curaçao) was mainly conducted in Papiamento

using face-to-face interview mode, while the DPHM2012 (the Netherlands) respondents’

interviews were mainly in Dutch and self-administered (digital or written). Differences in survey

language and administration modes potentially limit comparability between surveys and may have

introduced a bias in the prevalence estimates (21,22). The direction and magnitude of the bias may

differ per health variable. Since this may sometimes underestimate and other times overestimate

the differences we found among the study populations, we find it unlikely that this impacted our

overall conclusions.

Another limitation is that lower self-reported morbidity in Antillean non-migrants could reflect

underdiagnoses in this population. We, however, consider underdiagnoses unlikely to strongly

influence our results because the lower self-reports of disease are consistent with lower self-reports

of subjective health problems as limitations, psychological complaints and self-assessed health.

Interpretation

By combining the home and host country perspective on health disparities, our study shows new

opportunities for advancing the knowledge on migrant health and to differentiate whether

disparities between migrants and non-migrants are the result of biologically or culturally

determined differences, or the result of host country specific circumstances. Our study of Antillean

migrants in the Netherlands shows that the picture is diverse and depends on the health variable

studied.

Our finding that risk factor prevalence among Antillean migrants converge to the Dutch

population norm indicates that many Antilleans adopt health behaviors and attitudes of Dutch

society. This observation is in line with key-indicators of socio-cultural integration: Antillean

migrants frequently spend their free time with Dutch non-migrants and 45% of the marriages

among Antillean migrants is with a native Dutch partner (23). While cultural integration into the

host culture is generally regarded as the acculturation strategy with the best psychosocial outcomes

for migrant populations (24), the integration of a subgroup of Antillean migrants is reminiscent of

segmented assimilation. They represent a counterculture that is characterized by gang- and drug-

dictated lifestyles (25) that are associated with high drop-out rates (26), imprisonment rates (27),

and violent deaths (28).

The patterns of differences between the study populations for health status variables were

suggestive of the ‘stressful environment’ effect, with the exception of diabetes and overweight.

Chapter 2

44

The physical and mental health status was less favorable for Antillean migrant women than for

Antillean and Dutch non-migrant women. For Antillean migrant men, this was only observed for

limitations, morbidity and hypertension; rates for psychological complaints, migraine and self-

assessed health were similar to Dutch non-migrants, yet higher compared with Antillean non-

migrants. The results provide evidence against two potential underlying mechanisms of a ‘stressful

environment’ on Antillean migrants’ health. First, resettlement in a new country may be a stressful

experience that plausibly affects health, but this is unlikely to apply for Antillean migrants: we show

that their unfavorable health relative to the Dutch population persists in the second-generation

who were born in the Netherlands. Second, increased stress may lead to unhealthy coping

behaviors in the form of alcohol and tobacco use, which in turn lead to poorer health outcomes.

We tested this assumption by re-running the model for all health variables with the addition of risk

factor variables. This additional adjustment of our model did not attenuate the odds ratios of

Antillean migrants (Appendix table A5). In other words, our results imply that differences in

tobacco and alcohol use, overweight and physical inactivity do not explain differences in health

outcomes between the study populations.

Previous literature discusses three host-country specific environmental stressors that

disproportionally affect Antillean migrants more than the non-migrant population. First, almost

half (43%) of the Antillean migrants in the Netherlands report they have experienced negative or

unequal treatment based on their skin-color (29) and Dutch non-migrants and other migrant

groups were most negative in their judgement about Antillean migrants compared to other migrant

groups (30). Previous studies found that perceived discrimination and internalized racism –the

belief that one’s ethnic group is inferior caused by ethnic stereotyping- negatively affected physical

and mental health outcomes (31). Social stigmatization in Dutch society may therefore explain why

Antillean migrants’ health outcomes are less favorable compared to both non-migrant populations.

In fact, the migrant/ethnic/race variable that is traditionally used to capture cultural and biological

variations in health research, in this regard, may be a more accurate in capturing the consequences

of social stigmatization

Second, migrants are typically concentrated in urban areas (32). Health problems, such as drug-

and alcohol use, violent behavior, nutritional problems and mental health disease, and related social

difficulties are more prevalent in urban areas (33). While our results show that Antillean migrants

were twice as likely to live in an urban area than Dutch non-migrants (table 2), controlling for the

variable urbanization did not affect the overall patterns of significance (Appendix table A2 and

A3). This variable does, however, not reflect the spatial concentration of migrants within urban


45

areas in low socioeconomic, multi-ethnic neighborhoods (34), which creates a disproportionate

disadvantage in poor living conditions for migrant populations (35). Inhabitants of multi-ethnic

neighborhoods are more likely to report criminal victimization, reduced social cohesion and

feelings of unsafety and neighborhood deprivation than inhabitants of predominantly white

neighborhoods (30). This independent neighborhood effect may explain why Antillean migrants

in our study, as well as other migrants populations in the Netherlands (36), appear more susceptible

to hypertension and other stress-related disorders. Additionally, the concentration of migrants in

multi-ethnic neighborhoods also reduces interethnic contacts and possibly leads to a more negative

mutual perception of migrants and the non-migrant population (34).

Third, migrants in Western countries typically have lower levels of upward social mobility than

their non-migrant counterparts (37). This is also noticeable in the Netherlands: educational

achievement has been rising in all migrant groups relative to the Dutch non-migrants, but their

relative position in terms of unemployment, job insecurity, low income and poor housing has not

improved during the 2003-2015 period (23). Thus, at any given level of educational achievement,

migrants appear disadvantaged for factors that, directly or indirectly, contribute to good health.

This may explain why, despite being relatively better educated than their Dutch non-migrants’

counterparts (table 2), Antillean migrant women were more likely to report poor health outcomes.

Adjustment for educational status and employment, as we did in our analyses, may not fully control

for differences in the socioeconomic circumstances between Antillean migrants and Dutch non-

migrants.

The disadvantaged health outcomes of Antillean migrants in the Netherlands are likely the result

of complex interactions between their socioeconomic circumstances and their cumulative

experiences with environmental stressors during the life course (3). In part, this may also be

influenced by biological or cultural risk factors that persist after migration: our results show that

rates of diabetes (both genders) and overweight/obesity (men only) were similar in the two

Antillean populations. Additionally, vitamin D deficiency is more common among Antillean

migrants than among Dutch non-migrants due to an often higher pigmentation of the skin, which

may increase their susceptibility for certain physical and mental disorders (38).

Our finding that the poorer physical and mental health status of Antillean migrants is suggestive

of the ‘stressful environment’ effect is in line with previous literature that showed that prevalence

rates of schizophrenia in Afro-Caribbean’s are much higher in the United Kingdom than in

Jamaica, Trinidad and Barbados (39). Considering that the ‘host country’ perspective results of

migrant/ethnic health research are sometimes misinterpreted to validate ‘colonial’ stereotypes that

Chapter 2

46

potentially expand stigmatization and social differentiation (40), additional studies that take the

‘home country’ perspective into account may provide a valuable addition to inform policy makers.

The paradox may then not be found in the fact that some migrant groups can be considered

‘healthy migrants’, but that host country stressors still disproportionally affect migrants, and their

children, more than the non-migrant populations in some of the most developed countries in the

world.

Conclusions

Patterns of differences in physical and mental health among the three study populations were

suggestive of a ‘stressful environment’ effect. The poorer health of Antillean migrants in the

Netherlands is unlikely related to poorer health behaviors, but may be partly determined by host-

country specific stressors such as perceived discrimination, spatial concentration in multi-ethnic

neighborhoods and reduced social mobility. Future studies that take the home country perspective

of migrant health into account may provide a valuable addition to inform policy makers.

Acknowledgements

We gratefully acknowledge Dr. Albert Wong of the National Institute for Public Health and the

Environment (RIVM) for his valuable support with the statistical methods used in this paper.

Appendix

The Appendix for this study is available via this (private) link: www.vic.cw/appendices.


47

References

1. Razum O, Twardella D. Time travel with Oliver Twist--towards an explanation for a paradoxically low mortality among recent immigrants. Trop Med Int Health. 2002;7(1):4–10.

2. Angel JL, Buckley CJ, Sakamoto A. Duration or Disadvantage? Exploring Nativity, Ethnicity, and Health in Midlife. J Gerontol Soc Sci. 2001;56B(5):S275–84.

3. Spallek J, Zeeb H, Razum O. What do we have to know from migrants’ past exposures to understand their health status? A life course approach. Emerg Themes Epidemiol. 2011;8.

4. Abraído-Lanza AF, Dohrenwend BP, Ng-Mak DS, Turner JB. The Latino mortality paradox: a test of the “salmon bias” and healthy migrant hypotheses. Am J Public Health. 1999;89(10):1543–8.

5. Kennedy S, McDonald J, Biddle N. The healthy immigrant effect and immigrant selection: evidence from four countries. Soc Econ Dimens an Aging Popul Res Pap. 2006;(164).

6. Constant AF, García-muñoz T, Neuman S, Neuman T. A “Healthy Immigrant Effect” or a “Sick Immigrant Effect”? Selection and Policies Matter. Eur J Heal Econ. 2018;19(1):103–21.

7. Chiswick BR, Lee YL, Miller PW. Immigrant Selection Systems and Immigrant Health. Contemp Econ Policy. 2008;26(4):555–78.

8. Solé-Auró A, Crimmins EM. Health of immigrants in European countries. Int Migr Rev. 2008;42(4):861–76.

9. Rechel B, Mladovsky P, Deville W, Rijks B, Petrova-Benedict R, McKee M. Migration and health in the European Union. Rechel B, Mladovsky P, Deville A, Rijks B, Petrova-Benedict R, McKee M, editors. Maidenhead: Open University Press; 2011.

10. Ikram UZ, Kunst AE, Lamkaddem M, Stronks K. The disease burden across different ethnic groups in Amsterdam, the Netherlands, 2011-2030. Eur J Public Health. 2014;24(4):600–5.

11. Delavari M, Sonderlund A, Swinburn B, Mellor D, Renzaho A. Acculturation and obesity among migrant populations in high income countries - a systematic review. BMC Public Health. 2013;13(1):458.

12. Misra A, Ganda OP. Migration and its impact on adiposity and type 2 diabetes. Nutrition. 2007;23(9):696–708.

13. Veling W, Selten JP, Susser E, Laan W, Mackenbach JP, Hoek HW. Discrimination and the incidence of psychotic disorders among ethnic minorities in The Netherlands. Int J Epidemiol. 2007;36(4):761–8.

14. Boydell J, van Os J, McKenzie K, Allardyce J, Goel R, McCreadie RG, et al. Incidence of schizophrenia in ethnic minorities in London: ecological study into interactions with environment. BMJ. 2001;323(7325):1336–8.

15. Stirbu I, Kunst AE, Vlems FA, Visser O, Bos V, Deville W, et al. Cancer mortality rates among first and second generation migrants in the Netherlands: Convergence toward the rates of the native Dutch population. Int J Cancer. 2006;119(11):2665–72.

16. Harding S. Mortality of migrants from the Caribbean to England and Wales: Effect of duration of residence. Int J Epidemiol. 2004;33(2):382–6.

17. Volksgezondheidenzorg.info. Gezondheidsmonitor Volwassenen, GGD’en, CBS en RIVM [Internet]. [cited 2017 Aug 9]. Available from: https://bronnen.zorggegevens.nl/Bron?naam=Gezondheidsmonitor-Volwassenen%2C-

Chapter 2

48

GGD’en%2C-CBS-en-RIVM

18. Verstraeten SPA, Jansen I, Pin R, Brouwer W. De Nationale Gezondheidsenquête 2013: methodologie en belangrijkste resultaten. Willemstad, Curaçao: Volksgezondheid Instituut Curaçao; 2013.

19. Keij I. Standaarddefinitie allochtonen. Index No. 10. Den Haag: Centraal Bureau voor de Statistiek; 2000.

20. Berwick DM, Murphy JM, Goldman PA, Ware Jr. JE, Barsky AJ, Weinstein MC. Performance of a five-item mental health screening test. Med Care. 1991;29(2):169–76.

21. Heerwegh D, Loosveldt G. Face-to-Face versus Web Surveying in a High-Internet-Coverage Population: Differences in Response Quality. Public Opin Q. 2008 Dec 1;72(5):836–46.

22. Seo S, Chung S, Shumway M. How good is “very good”? Translation effect in the racial/ethnic variation in self-rated health status. Qual Life Res. 2014 Mar 13;23(2):593–600.

23. Huijnk W, Andriessen I. Integratie in zicht? De integratie van migranten in Nederland op acht terreinen nader bekeken. Den Haag: Sociaal en Cultureel Planbureau; 2016.

24. Berry JW. Acculturation: Living successfully in two cultures. Int J Intercult Relations. 2005;29(6 Spec. Iss.):697–712.

25. Oldenboom JFW. De Antilliaans-Nederlandse subcultuur van geweld. [Master thesis]: Erasmus University Rotterdam; 2009.

26. De Graaf W, Van Zenderen K. Segmented assimilation in the Netherlands? Young migrants and early school leaving. Ethn Racial Stud. 2009;32(8):1470–88.

27. Central Bureau of Statistics. CBS StatLine - Gedetineerden; geslacht, leeftijd en herkomstgroepering [Internet]. [cited 2017 Mar 15]. Available from: http://statline.cbs.nl/StatWeb/publication/?VW=T&DM=SLNL&PA=82321NED&LA=NL

28. Bos V, Kunst AE, Keij-Deerenberg IM, Garssen J, Mackenbach JP. Ethnic inequalities in age- and cause-specific mortality in The Netherlands. Int J Epidemiol. 2004;33(5):1112–9.

29. Andriessen I, Fernee H, Wittebrood K. Ervaren discriminatie in Nederland. Den Haag: Sociaal en Cultureel Planbureau; 2014.

30. Gijsberts M, Vervoort M, Havekes E, Dagevos J. Maakt de buurt verschil? De relatie tussen de etnische samenstelling van de buurt, interetnisch contact en wederzijdse beeldvorming. Den Haag: Sociaal en Cultureel Planbureau; 2010.

31. Pascoe EA, Richman LS. Perceived discrimination and health: A meta-analytic review. Psychol Bull. 2009;135(4):531–54.

32. Schiller NG, Caglar A. Locating Migration: Rescaling Cities and Migrants. 1st ed. Locating Migration Rescaling Cities and Migrants. Ithaca: Cornell University Press; 2011. 279 p.

33. Garretsen HF, Raat H. Urban health in the Netherlands: health situation, health facilities and public health policy. Health Policy. 1991 Jul;18(2):159–68.

34. Dagevos J. Ruimtelijke concentratie van niet-westerse migranten: achtergronden, gevolgen en aangrijpingspunten voor het beleid. Den Haag: Sociaal en Cultureel Planbureau; 2009.

35. Vlahov D, Freudenberg N, Proietti F, Ompad D, Quinn A, Nandi V, et al. Urban as a Determinant of Health. J Urban Heal. 2007;Suppl 1:16–26.


49

36. Agyemang C, C, Van Hooijdonk C, Wendel-Vos W, Ujcic-Voortman JK, Lindeman E, Stronks K, et al. Ethnic differences in the effect of environmental stressors on blood pressure and hypertension in the Netherlands. BMC Public Health. 2007;7(118).

37. Papademetriou DG, Somerville W, Sumption M. The Social Mobility of Immigrants and Their Children. Transatlantic Council on Migration. Washington DC: Migration Policy Institute; 2009.

38. Huibers MHW, Visser DH, Deckers MML, van Schoor NM, van Furth AM, Wolf BHM. Vitamin D deficiency among native Dutch and first- and second-generation non-Western immigrants. Eur J Pediatr. 2013;173(5):583–8.

39. Hickling FW. The epidemiology of schizophrenia and other common mental health disorders in the English-speaking Caribbean. Pan Am J public Heal. 2005;18(4–5):256–62.

40. Proctor A, Krumeich A, Meershoek A. Making a difference: The construction of ethnicity in HIV and STI epidemiological research by the Dutch National Institute for Public Health and the Environment (RIVM). Soc Sci Med. 2011;72(11):1838–45.

Chapter 3

Contribution of

amenable mortality to life expectancy differences between the constituent countries of the

Kingdom of the Netherlands

SPA Verstraeten, HAM van Oers and JP Mackenbach

Submitted

Chapter 3

52

Abstract

Objectives Despite being part of the same Kingdom, the Dutch Caribbean islands of Aruba

and Curaçao have substantially lower life expectancies than the Netherlands. In 2015, life

expectancy in Aruba was 6.7 years (men) and 5.4 years (women) shorter than in the Netherlands.

In Curaçao, this was 4.8 years and 2.4 years respectively. We examined the contribution of

mortality amenable to healthcare to the lower life expectancies in the Dutch Caribbean as

compared to the Netherlands.

Methods Analysis of age-standardized death rates and decomposition of life expectancy

differences between the Netherlands and the Caribbean islands of Aruba and Curaçao by age

and cause of death during the period 1988-2014.

Results We show that amenable mortality makes a substantial contribution to the life

expectancy gap with the Netherlands, particularly deaths from ischemic heart disease,

cerebrovascular disease, hypertensive disease, breast cancer, perinatal causes and

nephritis/nephrosis. If mortality from amenable causes were reduced to similar levels as in the

Netherlands, men and women in Aruba would add respectively 1.19 years and 0.72 years to

their life expectancies. In Curaçao, this would be 2.06 years and 2.33 years.

Conclusions Our study suggests that improvements in healthcare can help to reduce the gap

in life expectancy between the Dutch Caribbean and the Netherlands. In-depth studies

investigating care delivery related to these causes-of-death are necessary to identify the specific

interventions and resources needed for healthcare strengthening in the Dutch Caribbean.

Amenable mortality in constituent countries of the Kingdom of the Netherlands

53

Introduction

Life expectancies are higher in the Netherlands than in the Dutch Caribbean islands of Aruba and

Curaçao (1). In the Netherlands, life expectancy has increased steadily since 1985. Male life

expectancy in the Dutch Caribbean was two years lower in 1988, and failed to improve until the

late 2000s. This resulted in a substantial gap with the Netherlands of 6.7 years for Aruba and 4.8

years for Curaçao. Trends in female life expectancy, in contrast, moved more or less in parallel to

those of the Netherlands. In Aruba, this improvement was slower, thereby deteriorating its relative

position from 4.2 years in 1988 to 5.4 years in 2015. The female life expectancy gap somewhat

narrowed for Curaçao, from 2.7 years in 1988 to 2.4 years in 2015 (Supplementary Figure A1).

The determinants of these differences in life expectancy development are largely unknown and

require more specific explanations to inform policy strategies aimed at improving the health of the

Dutch Caribbean populations. Variations in life expectancies between populations are often the

result of complex interactions between many factors, e.g. socioeconomic conditions, behavioural

and environmental factors, and healthcare. Although the accessibility and quality of healthcare

services are not routinely monitored in the Dutch Caribbean, anecdotal evidence suggests that lack

of access to good quality healthcare services may play a role. Berend et al. reported on ten dialysis

patients in Curaçao who died due to acute aluminium intoxication caused by a defective water

distribution pipe (2). Alberts et al. showed that, despite much higher morbidity rates, the low-

educated on Curaçao were less likely to have consulted a specialist, and were also less likely to have

been hospitalized than the high-educated (3). Besides these studies, several country reports

describing aspects of healthcare are available, but their usefulness to inform policy development is

limited due to their fragmentary presentation. This lack of information demonstrates that

policymakers’ needs for relevant and timely information to identify issues, set priorities, support

practices and monitor progress in population health are not met in the Dutch Caribbean.

Moreover, there has been a lack of transparency in the findings of the islands’ healthcare

inspectorates, although the recent launch of departmental websites in Aruba (www.iva.aw) and

Curaçao (https://www.inspectiegmn.org/nl/) have laid a promising foundation for future

improvements.

The Dutch Caribbean islands have limited research capacity to routinely monitor aspects of health

and healthcare. This study is therefore an initial step to identify healthcare areas whose

optimization could contribute to reducing the gap in life expectancy with the Netherlands. The

results will enable the local governments to prioritize policy areas for more in-depth investigation,

with the ultimate goal of improving health planning. Using the concept of amenable mortality, our

Chapter 3

54

study aims to assess the contribution of healthcare to the lower life expectancies in the Dutch

Caribbean. Specifically, we examine trends in mortality amenable to health care during the 1988-

2014 period, and estimate the contribution of amenable causes-of-death to the gap in life

expectancy between the Dutch Caribbean islands and the Netherlands. Then, we look at national

health indicators and country reports to discuss the role of healthcare in our findings on amenable

mortality in more detail.

Methods

Data

We obtained mortality data from the World Health Organization (WHO) mortality database or

through a request at the Pan American Health Organization (PAHO) for Curaçao (4). Subject to

data availability, we included data of the Netherlands for the period 1988-2014, of Aruba for 1999-

2014 and of Curaçao for 1988-2007. Data include annual deaths by gender and age in each year,

using the tenth revision of the International Classification of Diseases (ICD). ICD-codes are given

in Supplementary Table A1 (and followed the classification of amenable deaths used in Nolte &

McKee (2004)) (5). Population data by gender and age were extracted from the United Nations

(UN) World Population Prospects (WPP) 2017 (1). We depicted data of the Netherlands Antilles

(which consisted of five island territories and was dissolved in 2010) and Curaçao (the most

populous island territory of the former Netherlands Antilles that accounted for 75% of the

population) as a continuous trend, since analyses showed a continuance of mortality trends from

the Netherlands Antilles (NA) (1988-2000) to Curaçao (2001-2007). We refer to these territories

as NA/Curaçao. Data sources for the comparison of health indicators between the Netherlands,

Aruba and NA/Curaçao are listed in the notes of Supplementary Table A4.

Data analyses

We corrected deaths for unclassified age and gender, for deaths attributed to ill-defined causes

(R00-R99, non-external deaths only) and for under-registration, as determined by comparing the

number of deaths with independent estimates from the WPP 2017 (6). Based on the corrected

numbers of deaths, we calculated age-standardized mortality rates per 100,000 population using

direct standardization to the World standard population (7). Changes in mortality during the study

period were determined with linear regression analysis (using joinpoint software, version 4.2.0.,

National Cancer Institute Bethesda, MD). We computed life expectancy from abridged life tables


55

and the age- and cause-specific contributions to life expectancy differences by Arriaga’s

decomposition method (8).

Results

Figure 1 demonstrates that the differences in amenable and non-amenable mortality between the

Netherlands and the Dutch Caribbean islands were substantial. From 1988 through 2014, both

amenable and non-amenable mortality has decreased steadily in the Netherlands, more rapidly for

men than for women. In Aruba and NA/Curaçao, trends have taken a more fluctuating course

due to their small populations, but levels of amenable and non-amenable mortality have been

consistently higher than in the Netherlands.

According to the results of the regression analysis (Supplementary Table A2), amenable mortality

has declined more rapidly than non-amenable mortality in all constituent countries, with the

exception of women in NA/Curaçao. Despite the strong fluctuations between 1999 and 2014, the

overall trend in amenable mortality in Aruba has been downwards. A comparison of the slopes

yielded non-significant results for both genders, i.e. the annual percentage changes were not

Figure 1 Trends in mortality from amenable and non-amenable causes, men and women, 0-74 year olds, Kingdom of the Netherlands, 1988-2014

Chapter 3

56

statistically significantly different between Aruba and the Netherlands. Consequently, the gap in

amenable mortality between Aruba and the Netherlands did not significantly widen or narrow

during the study period.

NA/Curaçao experienced a non-significant decline in amenable mortality rates during the 1988-

2007 period for both genders. A comparison of the slope with that of the Netherlands yielded

significant results for both genders, indicating that the annual percentage changes are different

between the Netherlands and NA/Curaçao. In contrast to the Netherlands, the NA/Curaçaon

population did not experience a decline of amenable mortality during the study period, so that the

gap in amenable mortality with the Netherlands widened.

Table 1 shows how much separate amenable causes-of-death contributed to the difference in life

expectancy between the Netherlands and the Dutch Caribbean islands during the latest time-

period for which data for all three constituent countries are available, i.e. 2005-2007. In Aruba,

amenable mortality accounts for 19% (men) and 17% (women) of the total mortality gap with the

Netherlands. For NA/Curaçao, this is substantially more with 37% and 75% respectively.

If mortality from amenable circulatory causes (which includes ischemic heart disease,

cerebrovascular disease, and hypertensive disease) were reduced to similar levels as in the

Netherlands, men and women in Aruba would add respectively 1.02 years and 0.56 years to their

life expectancies. In NA/Curaçao, this would be 0.65 years and 0.55 years. Perinatal deaths are an

important additional contributing cause-of-death in NA/Curaçao, and are responsible for 0.79

years (men) and 0.90 years (women) of the total mortality gap with the Netherlands.

Nephritis/nephrosis is also important, especially for NA/Curaçao. In addition, reduction in breast

cancer mortality to the level of the Netherlands would add 0.22 years in Aruba and 0.21 years in

NA/Curaçao to female life expectancy.


57

Table 1 Contribution of causes amenable to healthcare to the differences in life expectancy between the Netherlands and the Dutch Caribbean islands, Kingdom of the Netherlands, 2005-2007 Men Women Aruba NA/Curaçao Aruba NA/CuraçaoDifference in life expectancy (years) with the Netherlands due to: All causes of death 6.24 5.62 4.12 3.10 Non-amenable causes 5.05 3.57 3.41 0.77 Amenable causes 1.19 2.06 0.72 2.33 % of the difference due to amenable causes 19.1 36.6 17.4 75.1

Amenable causes Infectious diseases 0.01 0.01 0.01 0.00 Treatable cancers, among which: 0.05 0.04 0.18 0.35

- Colon and rectum 0.03 0.08 -0.08 0.10 - Breast -- -- 0.22 0.21 - Cervical and uterus -- -- 0.01 0.01 Diabetes 0.00 0.03 0.05 0.08 Ischemic heart disease (50% of deaths) 0.33 0.22 0.12 0.15

Cerebrovascular disease 0.34 0.20 0.32 0.20 Respiratory diseases 0.03 0.07 -0.06 0.06 Digestive diseases 0.00 -0.01 0.01 0.01 Perinatal deaths 0.09 0.79 -0.10 0.90 Other amenable conditions, among which: 0.34 0.40 0.19 0.58

- Hypertensive disease 0.35 0.23 0.12 0.20 - Nephritis and nephrosis 0.03 0.12 0.02 0.21

Discussion

Limitations

Before discussing our results from a substantive perspective we need to take the limitations of our

study into account. First, our analyses are constrained by the limited availability of mortality

statistics, by the uncertain accuracy and consistency of cause-of-death certification and coding, and

by the relatively small populations of the Dutch Caribbean. Preferably, mortality data would have

been available for the complete study-period in all constituent countries of the Kingdom of the

Netherlands, but this was not the case and is an important fact in itself, illustrating severe

shortcomings in health monitoring systems. Data of Aruba and NA/Curaçao also had more

inadequacies and were more incomplete than data from the Netherlands (Supplementary Table

A3). However, we corrected registered deaths for data inadequacies and incompleteness using

PAHO methods, so that the effects of data quality were minimized as much as possible (6). Due

to their small populations, the year-to-year fluctuations in mortality rates are relatively large in the

Chapter 3

58

Dutch Caribbean. This made it impossible to run joinpoint analyses for determining trend breaks

in the mortality series for the Dutch Caribbean.

Another limitation is that cross-country variation in amenable mortality does not automatically

imply that differences in healthcare performance are involved (9). Other factors that are not related

to healthcare performance, such as disease incidence, may also play a role in generating higher

mortality in the Dutch Caribbean. Disease incidence is also influenced by factors outside of the

healthcare system, such as socioeconomic conditions, health behaviours and environmental

factors. Many of these factors are less favourable in the Dutch Caribbean islands. For example,

income distributions are far more unequal than in the Netherlands and obesity is twice as prevalent

in Curaçao than in the Netherlands (Supplementary Table A4). Also, SO2-emissions recorded in

Curaçao are among the highest in the world (10). Observed differences in amenable mortality in

cross-country comparisons, however, cannot be solely attributed to differences in disease

incidence. In two recent studies, amenable mortality rates were controlled for incidence of

underlying conditions, yet substantial cross-country variations remained (11,12). Another

argument supporting the view that healthcare performance is involved in the difference in

amenable mortality between the Dutch Caribbean islands and the Netherlands is our finding that

amenable mortality has decreased faster than non-amenable mortality during the study period,

except for females in NA/Curaçao. This suggests that improvements in healthcare were at least

partly responsible for the mortality decline in amenable causes (13). This limitation also implies

that our results must be interpreted in light of other information on the healthcare system before

it is possible to conclude that the variation in amenable mortality rates reflect differences in the

effectiveness of healthcare services between Aruba, Curaçao and the Netherlands. Therefore, we

will discuss our findings in the light of available information from national health indicators and

country reports in the next section.

Interpretation

Deaths from causes that are considered unnecessary in the presence of timely and effective

healthcare substantially contribute to the life expectancy variations between the Netherlands and

the Caribbean islands of Aruba and NA/Curaçao. In Aruba, amenable deaths comprise 19% (men)

and 17% (women) of the total life expectancy difference with the Netherlands. In NA/Curaçao,

this is 37% and 75% respectively. The larger relative contribution of amenable deaths in

NA/Curaçao is mainly due to its relatively high mortality from perinatal causes and

nephritis/nephrosis. As the contribution in years is comparable between men and women in

NA/Curaçao, the large gender difference is explained by the relatively large contribution of non-


59

amenable mortality among males. Non-amenable mortality rates were also higher in the Dutch

Caribbean islands than in the Netherlands, especially for men, for whom the gap with the

Netherlands has strongly diverged during the study period. A previous study showed that the

divergence of mortality from non-amenable causes is largely due to mortality from homicide and

transport accidents (14).

All in all, amenable mortality plays a larger role in the life expectancy gap with the Netherlands for

NA/Curaçao than for Aruba. Ischemic heart disease, cerebrovascular disease, hypertensive disease

(hereafter referred to as circulatory diseases) and breast cancer are the underlying causes-of-death

with the largest contribution to life expectancy differences with the Netherlands. In NA/Curaçao,

perinatal deaths and nephritis/nephrosis also play a role. Medical interventions that help to prevent

deaths from these causes have been comprehensively described elsewhere and we briefly

summarize these insights here (15). Increased detection efforts for hypertension, improvements in

the treatment of hypertension from the 1950s onwards, and the intensive management of stroke

in coronary care units, have significantly reduced the mortality from circulatory diseases. Breast

cancer mortality declined after the introduction of population screening programs

(mammography) and treatment with tamoxifen, even though the effectiveness of the first

intervention remains a topic of dispute (16). Declines in deaths from perinatal causes have been

the result of advances in the treatment for specific conditions, such as rhesus immunization and

surfactant, and the incremental introduction of a wide range of interventions, such as special baby

care units, and local intensive and ventilator care. The key interventions in reducing deaths from

nephritis/nephrosis were dialysis and renal transplantation, and graft survival for the latter

treatment was further improved after the introduction of cyclosporine in the 1980s.

Accessibility

All healthcare systems in the Kingdom of the Netherlands are built on the principle of solidarity

and provide universal healthcare for their residents. In many aspects, the three healthcare systems

are comparable. For example, General Practitioners (GP) are the gatekeeper to healthcare; many

other health providers are only accessible upon referral by a GP. In practice, however, this seems

to work better in the Netherlands than in Curaçao (17). All countries’ healthcare systems are also

based on the value that everyone should receive the same type and quality of treatment, irrespective

of social status, gender and income. While equal access to the system generally has by-and-large

been achieved in the Netherlands (18), this was less the case in Curaçao, where greater needs for

services were not met by greater consumption (3). Part of the explanation was that waiting times

for specialist care in Curaçao were longer for low-income patients (whose insurance payment

Chapter 3

60

model was based on a capitation fee) than for private patients (whose insurance payment model

was based on fee-for-service) (19). With the introduction of the basic healthcare insurance system

in Curaçao in 2013, these differences in payment models and waiting times have largely

disappeared (20). Nevertheless, the results of a recent health survey show that the consumption of

specialist care was still not associated with self-reported morbidity (21). As access is commonly

defined as the use of healthcare by individuals with a need for care (22), this shows that access to

healthcare services is currently inadequate in Curaçao. On the one hand, this implies under

treatment of those who need medical services the most. On the other hand, this may also suggest

that overtreatment is common in Curaçao. Indeed, the consumption of primary health services is

much higher than in the Netherlands: more people had visited their general practitioner and had

used prescription medication, and twice as many drug prescriptions per inhabitant were recorded

(Supplementary Table A4). Unnecessary or inappropriate medical interventions are the largest

contributor of excess healthcare costs and an important cause of patient harm (23). We found no

comparable information on the accessibility and consumption of healthcare services for Aruba.

Quality of care

Universal healthcare does not automatically translate in lower amenable mortality, because this is

also dependent on the quality of the services (11). The legal framework that regulates the quality

of healthcare services clearly lags behind in the Dutch Caribbean. Thirteen years after the

introduction of the Individual Healthcare Professions Act (the so-called BIG-wet) in the

Netherlands, which regulates the qualification of healthcare professionals wanting to practice in

the healthcare sector, a similar law was proclaimed by the Curaçaon parliament (2010). The

provisions that stimulate quality standards of health providers and protects patients against

unskilful and careless behaviours have, however, not yet been implemented. Legislation that

regulates the quality of healthcare providers is not foreseen in Curaçao in the near future. Aruba

proclaimed a similar law as the BIG-wet, the AruBIG, in 2014. In the same year, Aruba also

introduced legislation that regulates the quality of healthcare providers. The provisions that these

regulations bring in force are currently implemented by Aruba’s healthcare inspectorate.

Our results show that circulatory diseases, breast cancer, perinatal deaths and nephritis/nephrosis

(the last two causes solely in NA/Curaçao) contribute to the lower life expectancies in the Dutch

Caribbean islands as compared to the Netherlands. There are several indications that patients with

these disorders and their underlying risk factors do not receive adequate curative and preventive

care in the Dutch Caribbean. During a health examination survey in Curaçao, blood pressure

control was 32% among respondents that were previously diagnosed with hypertension by a


61

medical professional (21). While data from the Netherlands were not directly comparable because

of different study methods, control of hypertension appears substantially higher among Dutch

men (53%) and women (61%) with hypertension in Amsterdam (24). Among diagnosed diabetics

in Curaçao, 35% had a normal blood glucose level (21), whereas 67% of diabetic men and 47% of

diabetic women in Amsterdam had HbA1c levels on target, which suggests that blood glucose

control is also less favourable in Curaçao (25). This demonstrates that the prevalence of

uncontrolled hypertension and diabetes, and subsequently the proportion of people that are at risk

for related complications such as nephritis/nephrosis, is relatively high in Curaçao. Information

on hypertension or blood glucose control in Aruba was not available.

Moreover, in order to prevent complications, it is recommended that patients with diabetes have

a regular eye and foot examination (Dutch College of General Practitioners (NHG) guidelines). In

2017, 67% of diabetics in Curaçao had undergone a clinical eye examination in the past 2 years,

and 29% had undergone a clinical foot-examination, a drop of 9% and 7% respectively compared

to 2013 (21). A severe complication of untreated diabetes is kidney disease. As another indication

that the detection and/or treatment of hypertension and diabetes is inadequate in Curaçao, the

proportion of dialysis patients was four times higher than in the Netherlands in 2014

(Supplementary Table A4). Concerns about an alarmingly high number of complications related

to diabetes were confirmed in Aruba as well (26). Further investigation should ascertain whether

the differences in nephritis/nephrosis mortality between the Dutch Caribbean islands is due to an

increased effectiveness of chronic disease care in Aruba that prevents the severe complications

from hypertension and diabetes, or is for example the result of different practices in the diagnosis

and treatment of renal failure.

A national breast screening program was introduced in the Netherlands in 1989. In NA/Curaçao,

a similar program was initiated by a private foundation in 2010, and only received financial support

from the local government in 2017. A national breast screening program was introduced in Aruba

in 2016. In 2017, breast cancer screening rates in Curaçao were lower than in the Netherlands,

which may partially explain the higher mortality-to-incidence ratios on the island. Information on

breast cancer screening rates were not available for Aruba. No information on the effectiveness of

breast cancer treatment in the Dutch Caribbean was available, as cancer survival rates are not

published by the local hospitals.

Curaçao also struggles with providing sufficient coverage in other areas of preventive care. For

cervix and colorectal cancer, relatively low screening rates and high mortality-to-incidence ratios

were recorded (Supplementary Table A4), which suggests room for further improvements in the

Chapter 3

62

diagnosis and/or treatment of these diseases. The association of GP’s in Curaçao recommends an

annual seasonal flu vaccination for every person at risk, but only 6% of the target population had

received one during the 2016 season (21). Moreover, the procedures necessary to quickly detect

and respond to infectious disease outbreaks are non-existent in Curaçao (27), and childhood

immunization coverage is below 95% (Supplementary Table A4). This is especially concerning in

light of the resurgence of vector-borne and vaccine-preventable outbreaks in infectious diseases

in the neighbouring country Venezuela (28,29). It is unclear whether the high coverage recorded

in Aruba indeed reflects higher population immunization rates, as the reported estimates preceding

the introduction of the child monitoring system from Curaçao also consistently depicted an overly

optimistic immunization coverage of >95%.

Perinatal deaths stem from healthcare-related causes such as the inappropriate management of

complications during pregnancy and delivery, and may be reduced by up to 30% through the

implementation of recommendations from quality-of-care audits (30). During the study period,

perinatal mortality was consistently higher in NA/Curaçao than in the other constituent countries

of the Kingdom of the Netherlands. Curaçao also recorded a maternal mortality rate that was three

times higher than in Aruba (Supplementary Table A4). This suggests that the effectiveness of

mother-and-child care is unsatisfactory in Curaçao. An important reason for Curaçao’s poorer

outcomes could be the fragmented organization of perinatal care services, with midwifes providing

services in the maternity clinic, and gynaecologists in the main hospital located several kilometres

away. In Aruba, in contrast, midwifes and gynaecologists work closely together in the main

hospital. Further investigation should determine whether the low perinatal mortality in Aruba

reflects lower incidences of underlying risk factors, better care during the antenatal period and

during delivery, and/or a registration artefact, for example from the exclusion of stillbirths.

All in all, the information on healthcare in the Dutch Caribbean suggests that the excess amenable

mortality from circulatory diseases, breast cancer, perinatal deaths and nephritis/nephrosis in the

Dutch Caribbean, at least partly, reflects differences in the effectiveness of healthcare services

between the Dutch Caribbean and the Netherlands. As we mentioned in the introduction, this

study is an initial step to identify healthcare areas whose optimization could substantially contribute

to reducing the gap in life expectancy with the Netherlands. Although our cross-country

comparison identifies areas where improvements are possible and necessary, they do not allow us

to specify which improvements are needed. More in-depth studies investigating care delivery

related to these amenable causes-of-death are therefore necessary, for example by linking these

findings to healthcare processes during consultations with policymakers, physicians and other


63

relevant health professionals. These efforts are expected to identify the specific interventions and

resources needed to reduce the mortality from these amenable causes in the Dutch Caribbean,

with the ultimate goal of improving health planning. During further investigations, the small scale

of the Dutch Caribbean islands should be taken into account. The volume of certain complex

specialist interventions, for example, may be too low to guarantee quality of care, which increases

the risks of mortality from the underlying conditions. To illustrate, surgeons in the Netherlands

commonly offer services in a specific area of specialization, while most surgeons in Aruba and

NA/Curaçao offer general services. Also, medical treatment abroad may be necessary because not

all medical specializations are available on the islands, which is likely to increase the waiting time

to start treatment for certain procedures. Nevertheless, every healthcare system, big or small, needs

to be optimized to provide the best care that the available resources can buy.

Conclusion

Our study suggests that improvements in the healthcare processes related to several amenable

causes-of-death can help to reduce the gap in life expectancy between the Dutch Caribbean and

the Netherlands. In-depth studies investigating care delivery related to these causes-of-death are

necessary to identify the specific interventions and resources needed for healthcare strengthening

in the Dutch Caribbean, with the ultimate goal of improving health planning.

Appendix


Chapter 3

64

References


2. Berend K, Van Der Voet G, Boer WH. Acute aluminum encephalopathy in a dialysis center caused by a cement mortar water distribution pipe. Kidney Int. 2001;59(2):746–53.

3. Alberts JF, Sanderman R, Eimers JM, Van Den Heuvel WJA. Socioeconomic inequity in health care: A study of services utilization in Curaçao. Soc Sci Med. 1997;45(2):213–20.


5. Nolte E, McKee M. Does health care save lives? Avoidable mortality revisited. London: The Nuffield Trust; 2004.

6. Silvi J. On the estimation of mortality rates for countries of the Americas. Epidemiol Bull Pan Am Heal Organ. 2003;24(4):1–5.

7. Ahmad OB, Boschi-Pinto C, Lopez AD, Murray CJL, Lozano R, Mie I. Age Standardization of Rates: A New WHO Standard. 2001. (GPE Discussion Paper Series). Report No.: 31.

8. Auger N, Feuillet P, Martel S, Lo E, Barry AD, Harper S. Mortality inequality in populations with equal life expectancy: Arriaga’s decomposition method in SAS, Stata, and Excel. Ann Epidemiol. 2014;24(8):575–80.

9. Mackenbach JP, Hoffmann R, Khoshaba B, Plug I, Rey G, Westerling R, et al. Using “amenable mortality” as indicator of healthcare effectiveness in international comparisons: results of a validation study. J Epidemiol Community Heal. 2013;67(2):139–46.

10. Pulster EL, Johnson G, Hollander D, McCluskey J, Harbison R. Exposure Assessment of Ambient Sulfur Dioxide Downwind of an Oil Refinery in Curaçao. J Environ Prot (Irvine, Calif). 2018;9(3):194–210.

11. Kruk ME, Gage AD, Joseph NT, Danaei G, García-Saisó S, Salomon JA. Mortality due to low-quality health systems in the universal health coverage era: a systematic analysis of amenable deaths in 137 countries. Lancet. 2018;392(10160):2203–12.

12. Alkire BC, Peters AW, Shrime MG, Meara JG. The economic consequences of mortality amenable to high-quality health care in low- and middle-income countries. Health Aff. 2018;37(6):988–96.

13. Mackenbach J, Bouvier-Colle M, Jougla E. “Avoidable” mortality and health services: a review of aggregate data studies. J Epidemiol Community Health. 1990;44(2):106–11.

14. Verstraeten SPA, van Oers HAM, Mackenbach JP. Differences in life expectancy between four Western countries and their Caribbean dependencies, 1980-2014. Eur J Public Health. 2019;pii: ckz10.

15. Plug I, Hoffmann R, Mackenbach JP, editors. Avoidable mortality in the European Union. Volume 1: Final report. Public Health. Rotterdam: EU Public Health Program 2007106; 2011.

16. Autier P, Boniol M, Koechlin A, Pizot C, Boniol M. Effectiveness of and overdiagnosis from mammography screening in the Netherlands: population based study. BMJ. 2017;359(j5224).

17. Snoeijs S, Boerma W. Evaluatie van de structuur van en de zorgverlening in de eerstelijnsgezondheidszorg op Curaçao. Utrecht: NIVEL; 2012.


65

18. Smits JPJM, Droomers M, Westert GP. Sociaal-economische status en toegankelijkheid van zorg in Nederland. RIVM rapport 279601002/2002. Bilthoven: Rijksinstituut voor de Volksgezondheid en Milieu (RIVM); 2002.

19. Bodok S, Kwidama E, Lourents N, Schroen M, Wijk van N, Kwidama L, et al. Wachten op specialistische zorg op Curaçao: betekent PP “Poko Poko”? IV. Willemstad, Curaçao: Eilandgebied Curaçao; 2008.

20. Verstraeten S. Resultaten monitor wachttijden medisch specialistische zorg. Willemstad, Curaçao: Volksgezondheid Instituut Curaçao; 2017.


22. Waters HR. Measuring equity in access to health care. Soc Sci Med. 2000;51(4):599–612.

23. Lyu H, Xu T, Brotman D, Mayer-Blackwell B, Cooper M, Daniel M, et al. Overtreatment in the United States. PLoS One. 2017;

24. Agyemang C, Kieft S, Snijder MB, Beune EJ, Van Den Born BJ, Brewster LM, et al. Hypertension control in a large multi-ethnic cohort in Amsterdam, The Netherlands: The HELIUS study. Int J Cardiol. 2015;183:180–9.

25. Snijder MB, Agyemang C, Peters RJ, Stronks K, Ujcic-Voortman JK, van Valkengoed IGM. Case Finding and Medical Treatment of Type 2 Diabetes among Different Ethnic Minority Groups: The HELIUS Study. J Diabetes Res. 2017;2017:1–8.

26. Coenen T. Making a difference through health. How PwC is Helping to Change Lives. Impact case studies. Oranjestad, Aruba: PricewaterhouseCoopers; 2016.

27. Algemene Rekenkamer Curaçao. Chikungunya, aanpak van de bestrijding. Willemstad: Algemene Rekenkamer Curaçao; 2016.

28. Paniz-Mondolfi AE, Tami A, Grillet ME, Márquez M, Hernández-Villena J, Escalona-Rodríguez MA, et al. Resurgence of vaccine-preventable diseases in Venezuela as a regional public health threat in the Americas. Emerg Infect Dis. 2019;25(4):625–32.

29. Grillet ME, Hernández-Villena J V, Llewellyn MS, Paniz-Mondolfi AE, Tami A, Vincenti-Gonzalez MF, et al. Venezuela’s humanitarian crisis, resurgence of vector-borne diseases, and implications for spillover in the region. Lancet Infect Dis. 2019;3099(18):1–13.

30. Pattinson R, Kerber K, Waiswa P, Day LT, Mussell F, Asiruddin S, et al. Perinatal mortality audit: Counting, accountability, and overcoming challenges in scaling up in low- and middle-income countries. Int J Gynecol Obstet. 2009;107(suppl):S113–21.

Part II The political context of health and health policy

performance in the Caribbean

Chapter 4

Decolonization and life expectancy in the Caribbean


Social Science & Medicine 2016, vol. 170, pp. 87-96

https://doi.org/10.1016/j.socscimed.2016.08.048

Chapter 4

70

Abstract

Introduction Decolonization has brought political independence to half the Caribbean states

in the last half of the 20th century, while the other states remain affiliated. Previous studies

suggested a beneficial impact of affiliated status on population health, which may be mediated

by more favorable economic development. We assessed how disparities in life expectancy

between currently sovereign and affiliated states developed over time, whether decolonization

coincided with changes in life expectancy, and whether decolonization coincided with similar

changes in GDP per capita.

Methods Time-series data on life expectancy and related variables, GDP per capita and

political status were collected from harmonized databases. We quantified variations in life

expectancy by current political status during the 1950-2010 period. We assessed whether

decolonization coincided with life expectancy trend changes by: 1. calculating the annual

changes before and after independence, and 2. evaluating trend breaks in a predefined period

during decolonization using joinpoint analyses. Similar analyses were undertaken for GDP per

capita.

Results Life expectancy in currently sovereign Caribbean states was already lower than in

affiliated states before political independence. Overall, decolonization coincided with reductions

in life expectancy growth, but not with reductions in economic growth, and changes in life

expectancy growth in the decade after independence did not correspond with changes in

economic performance. The widening of the life expectancy gap between currently sovereign

and affiliated states accelerated in the 1990’s and continues to increase.

Conclusions Despite considerable life expectancy gains in all Caribbean states, life expectancy

in currently sovereign states increasingly lags behind that of states which remained affiliated.

Our results indicate that changing economic conditions were not the main determinant of the

unfavorable trends in life expectancy during and after decolonization. Circumstantial evidence

points to the weakening of bureaucracies during decolonization underlying the uneven life

expectancy developments of currently affiliated and sovereign states.


71

Introduction

Decolonization has brought political independence to half the Caribbean states in the last half of

the 20th century, while the other states remain affiliated to their former imperial countries. A recent

study demonstrated that since 1965, life expectancy growth has been remarkably more uneven

among Caribbean states than in the rest of the Americas (1). As a consequence, the Caribbean

currently has the lowest life expectancies and highest between-country disparities of the Americas.

Historical evidence reveals that a multitude of factors has contributed to the global rise in life

expectancy, including medical and public health interventions, rising levels of education, economic

development and income redistribution, and improving social conditions. Many of these factors

are influenced by political decision-making, and empirical evidence of the impact of politics

(defined as the process of making decisions applying to all members of a population) on population

health is growing rapidly (2–4). Politics may directly shape the conditions for life expectancy

improvement through policies related to primary and secondary prevention programs, road traffic

safety, food, water and air safety, health promotion campaigns, and health care delivery, or

indirectly through policies related to education, social and economic development (5,6) and

policies that address income inequalities (7).

Decolonization - the process in which a colony becomes politically independent - is one important

political event that can plausibly affect population health. Economic analyses that exploit current

differences in political status have found that affiliated status is consistently associated with better

economic conditions and better health outcomes (8,9). These studies remain inconclusive,

however, on the onset of health disparities between currently sovereign and affiliated states, and

studies that directly investigate the impact of decolonization on population health remain scarce.

While improved health outcomes were reported in several former colonies after political

independence (10–12), population health in many former colonies remains relatively poor to the

present day.

Decolonization may affect life expectancy development in several ways. First, decolonization can

lead to lower economic growth relative to other independent countries (13), for example through

discouraging private investments (14). Given the strong dependency of life expectancy growth on

economic conditions (15), even in the modern age (16), this may affect life expectancy

developments during decolonization. If this were the case, we expect that life expectancy growth

during decolonization mirrors a state’s economic growth. Second, decolonization can be

accompanied by civil unrest and violent eruptions. Short-term effects on life expectancy growth

Chapter 4

72

include an increase in the number of violent deaths, the displacement of civilians, the disruption

of health services and economic activities, and the reallocation of government resources from

public services to military expenditures (17). The effects of civil unrest on population health persist

long-term and result in additional health challenges, for example through increases in infectious

and other preventable diseases (18) and mental health problems (17). Third, political decisions

during decolonization can lead to the weakening of (colonial) bureaucracies and, consequently, a

state’s administrative capacity (19). In the absence of strong bureaucracies, the government’s

capacity to solve problems, to provide necessary resources and to implement effective policy

initiatives is challenged in areas that contribute to population health, such as health care, education

and social welfare. Political decisions during decolonization were also found to influence the

termination of beneficial medical reforms (20) and to incite health professional migration to

Western countries (21).

The purpose of this paper is to explore the development of population health -indicated by life

expectancy at birth- and the potential role of changing economic conditions -indicated by GDP

per capita- during decolonization in the Caribbean region in the last half of the 20st century.

Specifically, we aim to 1) quantify variations in life expectancy development by political status

during the 1950-2010 period, 2) assess whether decolonization coincided with changes in life

expectancy growth, and 3) assess whether decolonization coincided with similar changes in GDP

per capita.

Methods

Data

We selected datasets that were most comprehensive in terms of study period and number of

included Caribbean states based on searches in harmonized databases. The source of life

expectancy at birth data is the World Population Prospects 2012 dataset of the United Nations

(UN), which included states with 90,000 inhabitants or more. We collected life expectancy data in

single calendar years (interpolated from 5-year estimates), by gender, from 1950 to 2009 (22). In

order to provide explanatory clues and to confirm our findings, we selected three variables from

the same dataset that are related to life expectancy: crude death rate (per 1,000 population, by

gender), under-5 mortality and infant mortality (both per 1,000 live births). The source of GDP

per capita data is the Bulmer-Thomas dataset (in US$, 2000 prices), which includes annual time

series on 28 Caribbean states, from 1960 to 2008 (23).


73

Definition and selection of Caribbean states and their political status

The Pan American Health Organization’s (PAHO) definition of the Caribbean includes 30 states:

27 island states and 3 continental states in South America: French Guiana, Suriname and Guyana.

Continental Belize in Central America was added to this selection because of its strong historical

and economic ties to the Caribbean and its membership to the Caribbean Community and

Common Market (CARICOM). Life expectancy data of 21 states were available in the UN dataset.

Information on political status, including the year of independence where relevant, was collected

from the United Nations (http://www.un.org/en/decolonization/nonselfgov.shtml#s) and

verified with other sources, such as national websites. We included sovereign states -loosely

defined as populations in defined territories that are represented by one government and have full

control of their (inter)national affairs by international law- that gained political independence

during the period for which life expectancy data were available. This excludes Haiti, the Dominican

Republic and Cuba. Table 1 presents a factual overview of the political status of the 18 states

included in this analysis during the 1950-2010 period.

The number of independent Caribbean states grew from three in 1960 to sixteen in 1983 during

three consecutive waves of decolonization. The majority of these states were former British

colonies and all are currently member of the Commonwealth of Nations, an a-political

intergovernmental organization with the purpose of collaborating on shared objectives. Surinam

is the only former Dutch colony that gained political independence. All currently sovereign states

are sovereign under international law and UN member states.

Within the same time period other imperial countries with presence in the Caribbean, the

Netherlands, France and the United States (U.S.), established new political entities for their

colonies. The U.S. Virgin Islands is an unincorporated territory with limited political power and

citizenship rights. Aruba, Curaçao and Puerto Rico, have gained more autonomy and are self-

governing with respect to internal affairs. The three French colonies, French Guiana, Guadeloupe

and Martinique, became French departments in 1947. Its citizens have full citizenship rights and

the states are considered an integral part of the French Republic. All currently affiliated states have

strong political, legal and (socio)-economic ties to their former imperial power.

Cha

pter

4

74

Tab

le 1

Cha

nge

in p

oliti

cal s

tatu

s of C

arib

bean

stat

es in

clude

d in

this

analy

sis, 1

950-

2010

Ca.

195

0 C

a. 1

955

Ca.

196

0 C

a. 1

965

Ca.

197

0 C

a. 1

975

Ca.

198

0 C

a. 1

985

Ca.

199

0 C

a. 1

995

Ca.

200

0 C

a. 2

005

Ca.

201

0

Cu

rren

tly

sove

reig

n s

tate

sA

ntig

ua a

nd

Barb

udaa

U

K

UK

U

K

UK

U

K

UK

U

K

Inde

p '81

In

dep

Inde

p In

dep

Inde

p In

dep

Baha

mas

a U

K

UK

U

K

UK

U

K

Inde

p '73

In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Barb

ados

a U

K

UK

U

K

UK

In

dep

'66

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Beliz

ea,b

U

K

UK

U

K

UK

U

K

UK

'73

UK

In

dep

'81

Inde

p In

dep

Inde

p In

dep

Inde

p G

rena

daa

UK

U

K

UK

U

K

UK

In

dep

'74

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p G

uyan

aa

UK

U

K

UK

U

K

Inde

p '66

In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p Ja

maic

aa

UK

U

K

UK

In

dep

'62

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p St

. Luc

iaa

UK

U

K

UK

U

K

UK

U

K

Inde

p '79

In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p St

. Vin

cent

& th

e G

rena

dine

sa U

K

UK

U

K

UK

U

K

UK

In

dep

'79

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Surin

ame

Net

her-

lands

CC

KN

'54

CC

KN

CC

KN

CC

KN

In

dep

'75

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p

Trin

idad

& T

obag

oa

UK

U

K

UK

In

dep

'62

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p In

dep

Inde

p

Cu

rren

tly

affi

liate

d s

tate

s

Aru

ba

Net

her-

lands

N

A '5

4 N

A

NA

N

A

NA

N

A

NA

CC

KN

'86

CC

KN

CC

KN

CC

KN

CC

KN

Cura

çao

Net

her-

lands

N

A '5

4 N

A

NA

N

A

NA

N

A

NA

N

A

NA

N

A

NA

CC

KN

' 10

Fr

ench

Gui

ana

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Fran

ce

Gua

delo

upec

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

M

artin

ique

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Fr

ance

Pu

erto

Rico

U

S U

S U

S U

S U

S U

S U

S U

S U

S U

S U

S U

S U

S U

.S. V

irgin

Islan

ds

US

US

US

US

US

US

US

US

US

US

US

US

US

Not

es: a

Inde

pend

ent s

tate

with

in th

e Co

mm

onw

ealth

of N

atio

ns, b

Nam

e w

as c

hang

ed fr

om B

ritish

Hon

dura

s to

Beliz

e in

197

3, c

Life

exp

ecta

ncy

data

incl

udes

Sain

t-Bar

thél

emy

and

Sain

t-Mar

tin (F

renc

h pa

rt).

Abb

revi

atio

ns: C

CKN

: Con

stitu

ent c

ount

ry w

ithin

Kin

gdom

of t

he N

ethe

rland

s (‘ye

ar),

Inde

p: In

depe

nden

t sta

te (‘

year

), N

A: N

ethe

rland

s Ant

illes

, a c

onst

ituen

t cou

ntry

with

in

Kin

gdom

of t

he N

ethe

rland

s, U

K: c

olon

y of

the

Briti

sh E

mpi

re, U

S: u

ninc

orpo

rate

d te

rrito

ry o

f the

Uni

ted

Stat

es.


75

Quantitative analysis

To conclude whether decolonization coincides with a changes in population health or GDP per

capita, two quantitative approaches were applied that together would offer consistent evidence

when consensus in outcomes was found. The analyses were applied to all available variables, for

each currently sovereign state.

Growth developments before and after independence

Absolute and relative annual growth was calculated in the ten-year period before and after the year

of independence. Measured in years or US$, the absolute growth is the arithmetic difference

between life expectancy and GDP per capita respectively at the year of independence and ten years

before or after independence. Relative annual growth was calculated as the median annual

percentage of growth during the ten years before and after independence. Statistical significance

of differences in life expectancy growth before and after independence was assessed with the one-

sided Wilcoxon Signed-rank test. Strength of the association in life expectancy growth before and

after independence was tested with the Spearman’s Rank Correlation Coefficient (rho).

Joinpoint analysis

Trend changes during the period for which time-series data were available were determined with

joinpoint regression software (version 4.2.0., National Cancer Institute Bethesda, MD). For this

analysis, a minimum number of 0 and a maximum number of 3 joinpoints were supplied and

homoscedasticy was assumed. In order to narrow the time period within which a trend change in

mortality and economic conditions could be expected as a result of the transition to independence,

we chose a 6-year period; the year of independence, 1-year before and 4-years after.

The results of the joinpoint analysis were analyzed in line of previous research (24):

Count of matches: Measured in total number of matches, it counts the number of joinpoints

occurring in the predefined period of 1-year before and 4-years after the year of independence.

When more than half of the states show a match during this period, this was considered evidence

consistent with a change in mortality and economic conditions coinciding with decolonization.

Likelihood test: Under the assumption that joinpoints occur randomly during the 60-year (for

health variables) or 48-year (for GDP per capita) study period, the theoretical probability of a

match in a predefined period of 1-year before and 4-years after the year of independence was

calculated. From this theoretical probability the expected number of matches was derived. The

Chapter 4

76

expected number of matches was compared with the observed number of matches using Fisher’s

exact test. When the number of observed matches was significantly higher than expected, this was

considered evidence consistent with a change in mortality and economic conditions coinciding

with decolonization.

Since the selection of the predefined 6-year time period (year of independence, 1-year before and

4-years after) is somewhat arbitrary, we also calculated the number of matches occurring in 6-year

time periods set earlier or later around the year of independence. We found that the overall

conclusions of the joinpoint analysis did not change when the 6-year time period was set later (i.e.,

year of independence and 5 years after). However, when the 6-year time period was set earlier

(beginning 2, 3 or 4 years before independence), the results of Fisher’s exact test became non-

significant, indicating that the number of matches found in the joinpoint analysis was no longer

significantly higher than would be the case by pure chance alone (Appendix table A1). This shows

that our conclusions are sensitive to the time period chosen for the analysis, but as we consider it

unlikely that sudden life expectancy changes occurring before decolonization are causally related

to decolonization, we consider our findings sufficiently robust.

Results

Secular trends in life expectancy in Caribbean states, 1950-2010

Table 2 shows that life expectancy has risen steadily in currently sovereign states, but not as fast

as in currently affiliated states. In 1950, Caribbean states that would become independent in the

decades thereafter already lagged behind by 2.9 years (men) and 1.9 years (women), as compared

with the median life expectancy in currently affiliated states. When the eleven former colonies

became politically independent between 1962 and 1981, overall life expectancy remained relatively

low (more for women than for men), but stable during the entire period. In the 1990’s, there was

a period of stagnation that affected currently sovereign states more than affiliated states. As a

consequence, the gap between currently sovereign and affiliated states increased further from 1990

to 2009 with an additional 1.1 years for men and 2.0 years for women. By 2009, male life expectancy

in currently sovereign states lagged behind by four years as compared with currently affiliated

states, making it similar to the level of life expectancy in currently affiliated states in 1984. Female

life expectancy in currently sovereign states was seven years lower in 2009, making it comparable

to the level observed in affiliated states in 1983.


77

To explore life expectancy developments in individual sovereign states, we compared their life

expectancy development with the median of currently affiliated states. Figure 1 illustrates when

trend changes in male life expectancy occurred before, during and after independence during the

three waves of decolonization. In 1950, male life expectancy was higher in Trinidad & Tobago and

the Bahamas compared with the median of currently affiliated states. The onset of divergence from

the currently affiliated states showed much variety among states. In Jamaica and Trinidad &

Tobago, the first two states that gained independence in 1962, an abrupt negative trend change

coincided with the transition to independence. In Guyana, the Bahamas, Grenada and Surinam,

the gap with affiliated states was already present before political independence. The decades after

independence witnessed an increased widening of the gap caused by a (further) stagnation of life

expectancy growth or, in the case of Belize and Jamaica, a decrease in life expectancy. At the end

of the study period life expectancy was lower in all currently sovereign states, ranging from -1.38

years in Antigua & Barbuda to -11.21 years in Guyana. Only three states were able to improve

their relative position during independence as compared with the currently affiliated states:

Barbados, Antigua & Barbuda and St. Lucia.

Female life expectancy developments in currently sovereign states are similar, yet more

pronounced (Appendix figure A1). Female life expectancy was higher in Antigua & Barbuda, the

Bahamas and Guyana in 1950, compared with the median of the currently affiliated states. In 2009,

however, life expectancy was lower in all currently sovereign states, including Antigua & Barbuda,

the Bahamas and Guyana. Differences with the affiliated states ranged from -4.38 years in Antigua

& Barbuda to -13.68 years in Guyana. None of the currently sovereign states was able to improve

their relative position to the currently affiliated states during independence. In other words, despite

a small convergence observed in the 1980’s, female life expectancy in all currently sovereign states

increasingly lagged behind.

Cha

pter

4

78

Tab

le 2

Life

exp

ecta

ncy

in th

e Ca

ribbe

an st

ates

from

195

0 to

201

0, b

y ge

nder

and

cur

rent

pol

itica

l sta

tus

L

ife

exp

ecta

ncy

(in

yea

rs)

Cu

rren

t p

olit

ical

st

atu

s

# o

f st

ates

19

50-

1954

19

55-

1959

19

60-

1964

19

65-

1969

19

70-

1974

19

75-

1979

19

80-

1984

19

85-

1989

19

90-

1994

19

95-

1999

20

00-

2004

20

05-

2009

Men

Sove

reig

n

11

54.5

57

.0

59.5

61

.6

63.4

64

.3

65.9

67

.3

68.0

67

.5

68.4

69

.6

Q1-

Q3

54.2

-56.

2 55

.8-5

8.8

58.2

-60.

8 60

.5-6

2.7

61.9

-63.

8 63

.6-6

5.2

64.6

-66.

6 65

.3-6

7.9

65.9

-68.

7 66

.1-6

8.9

66.5

-69.

9 67

.8-7

1.2

Aff

iliat

ed

7 57

.4

60.5

62

.2

63.9

65

.6

67.4

69

.3

70.4

71

.1

71.3

72

.3

73.8

Q

1-Q

3 52

.6-5

8.7

55.9

-62.

6 59

.0-6

4.7

61.8

-66.

2 64

.2-6

7.5

66.3

-69.

0 68

.0-7

0.4

69.7

-70.

9 70

.4-7

1.5

71.2

-72.

9 71

.7-7

4.5

72.5

-75.

7

Dif

fere

nce

(y

ears

)

2.9

3.5

2.7

2.3

2.3

3.1

3.4

3.2

3.1

3.8

3.9

4.2

Wom

en

So

vere

ign

11

58

.4

60.8

62

.6

65.5

67

.8

69.6

71

.1

72.5

73

.1

73.1

73

.8

74.8

Q

1-Q

3 57

.3-5

9.8

60.0

-62.

4 62

.4-6

5.2

64.7

-66.

9 66

.9-6

8.4

68.3

-70.

5 69

.8-7

2.2

70.7

-73.

2 71

.2-7

3.5

72.0

-74.

1 72

.8-7

5.2

73.4

-76.

5 A

ffili

ated

7

60.3

64

.2

66.8

69

.2

71.7

73

.8

75.2

76

.2

77.9

79

.4

80.7

81

.7

Q1-

Q3

56.6

-61.

6 59

.6-6

5.3

62.9

-68.

0 66

.7-6

9.9

69.2

-72.

2 71

.5-7

4.3

73.9

-75.

9 75

.9-7

7.4

76.8

-78.

5 77

.6-8

0.0

78.7

-81.

2 79

.6-8

2.3

Dif

fere

nce

(y

ears

)

1.9

3.4

4.2

3.6

3.9

4.2

4.0

3.8

4.9

6.3

6.9

6.9

Dat

a re

fer t

o th

e sa

me

sele

ctio

n of

stat

es d

urin

g th

e st

udy-

perio

d, n

amel

y th

e cu

rren

tly 1

1 so

vere

ign

or 7

aff

iliat

ed st

ates

A

bbre

viat

ions

: Q1=

first

qua

rtile

, Q3=

third

qua

rtile


79

Figure 1 Differences in life expectancy between currently sovereign and affiliated states 1950-2010, men

Chapter 4

80

Quantitative analysis of life expectancy developments during decolonization

The observation that life expectancies between currently sovereign and affiliated states increasingly

diverged after independence in sovereign states is confirmed in the quantitative analyses. Appendix

table A2 shows the results of the first quantitative approach. Life expectancy growth was smaller

in the decade after than in the decade before independence for both genders, except in St. Vincent

& the Grenadines and Guyana. This corresponded with a decreased decline in crude death rate,

which was -1.5 for men and -1.6 for women before independence, but lowered to -0.5 for men

and -0.7 for women after independence and became positive in Antigua & Barbuda (Appendix

table A3 and A4). Only in St. Vincent & the Grenadines, the male crude death rate dropped faster

after independence, corresponding with an increased growth of male life expectancy. The changes

in mortality rates were not the same for all ages across currently sovereign states. The relative

annual growth for infant and under-5 mortality remained similar in five states after their

independence (Appendix table A5 and A6). The relative annual growth rate for infant mortality

was, in contrast to under-5 mortality, not significantly different before and after independence

when all states were considered.

Considering the currently sovereign states individually, differences in life expectancy growth were

most dramatic in Belize and Trinidad & Tobago, where growth in the decade after independence

declined with approximately 80 per cent for men and 70 per cent for women. In Belize, despite its

favorable starting position at the time of independence, male and female life expectancy started to

decrease in the mid 1980’s. In Surinam, a strongly negative development in infant mortality was

seen. Relative annual growth was -2.1% before independence, but came to a complete halt in the

following decade (Appendix table A6).

Table 3 summarizes the results of the second quantitative approach, showing the estimated

calendar years in which a joinpoint (of potentially three) occurred that was closest to the year of

independence, and whether or not this can be considered a match. A match was defined as the

occurrence of a joinpoint that indicated a negative trend change during the predefined 6-year time

period. A match was found in six of eleven states that became sovereign during the study period:

for both genders in Guyana, Jamaica and Trinidad & Tobago, for men only in Surinam and for

women only in Belize and St. Lucia. A corresponding match indicating a negative development in

crude death rate was found for Guyana, Jamaica and Trinidad & Tobago, for infant mortality in

Surinam and Trinidad & Tobago and for under-5 mortality in Jamaica, St. Lucia and Trinidad &

Tobago (Appendix table A7). Only in Belize the negative trend change in female life expectancy

could not be confirmed with corresponding trend changes in the three mortality variables. Table


81

4 shows the results of the likelihood test, in which we determined whether the number of matches

found was significantly higher than would be the case by pure chance. The test indicates that the

assumption of random distribution can be rejected.

Table 3 Results joinpoint analysis and count of matches

Year of Indep

Gender Year of closest

joinpoint (95% CI) Change in

APCa Closest trend change (years)

Match (1 year before, 4 years

after indep) Antigua & Barbuda

1981 M 1988 (1984-1993) -0.1 7 No F 1975 (1971-1979) -0.1 -6 No

Bahamas 1973 M 1978 (1974-1982) -0.1 5 No F 1966 (1963-1968) -0.3 -7 No

Barbados 1966 M 1962 (1959-1964) -0.3 -4 No F 1962 (1959-1965) -0.2 -4 No

Belize 1981 M 1987 (1985-1988) -0.9 6 No F 1985 (1984-1987) -0.6 4 Yes

Grenada 1974 M 1980 (1978-1984) -0.1 6 No F 1969 (1966-1973) -0.2 -5 No

Guyana 1966 M 1969 (1959-1984) <-0.1 3 Yes F 1968 (1965-1970) -0.1 2 Yes

Jamaica 1962 M 1962 (1960-1964) -0.9 0 Yes F 1961 (1960-1963) -0.9 -1 Yes

St. Lucia 1979 M 1984 (1982-1985) -0.7 5 No F 1982 (1981-1984) -0.8 3 Yes

St. Vincent & the Grenadines

1979 M 1989 (1986-1992) -0.3 10 No F 1987 (1985-1990) -0.4 8 No

Suriname 1975 M 1979 (1976-1982) -0.3 4 Yes F 1984 (1983-1986) -0.4 9 No

Trinidad & Tobago

1962 M 1963 (1962-1964) -0.6 1 Yes F 1962 (1961-1982) -0.9 0 Yes

aAPC is the Annual Percentage Change in the years before or after a joinpoint occurred. Depicted is the difference in APC before and after the joinpoint closest to year of independence occurred. The negative number therefore indicates decreased life expectancy growth.

Chapter 4

82

Table 4 Results likelihood test Length of observation period 60 years Length of period expected effect 6 years Theoretical probability of a match 0.10 Maximum number of matches (both genders) 22 Expected number of matches 2 Observed number of matches 9 Observed probability of a match 0.41 P-value Fisher’s exact test (one-sided) 0.017 Under the assumption that joinpoints occur randomly during the 60-year study-period, the theoretical probability of a joinpoint occurring in the predefined period of 1-year before and 4-years after the year of independence is 0.10. Of 22 possible matches (11 states, 2 genders), the expected number of matches is 2. The observed number of matches was significantly higher than would be the case by chance.

The role of economic development during decolonization

Similar analyses were undertaken for GDP per capita (in US$, 2000 prices) to explore changing

economic conditions during decolonization. In 1960, Caribbean states that would become

independent in the decades thereafter already lagged behind by US$2,735 as compared with the

median GDP per capita in currently affiliated states, making the GDP per capita three times higher

in affiliated states. After decolonization between 1962 and 1981, GDP per capita continued to

diverge due to stagnation of economic growth or, in the case of the Bahamas, a decline in absolute

GDP per capita. This totaled a US$11,750 or 256% difference in GDP per capita in the period

2005-2008 (Appendix table A8).

While the median GDP per capita in currently sovereign states was lower compared with the

affiliated states during the whole study period, the relative annual growth was not significantly

lower after independence (4.5% vs. 3.5%) (Appendix table A9). In addition, the results of the count

of matches and likelihood test were non-significant (Appendix table A10 and A11). Considering

the currently sovereign states separately, a decline in economic growth was apparent in seven of

eleven Caribbean states in the decade after independence. This decline was especially pronounced

in the high-income state of the Bahamas, where growth of GDP per capita and life expectancy

both steeply declined during the transition to independence. In contrast to the life expectancy

decline, decline in GDP per capita did not coincide with decolonization (Appendix table A10) but

started 4 years before the year of independence. In Antigua & Barbuda, Guyana and Jamaica, GDP

per capita more than doubled in the decade after independence while these countries experienced

a stagnation of life expectancy growth.


83

Discussion


Life expectancy in currently sovereign Caribbean states was already lower than in affiliated states

before political independence. Overall, decolonization coincided with reductions in life expectancy

growth, but not with reductions in economic growth, and changes in life expectancy growth in the

decade after independence did not correspond with changes in economic performance. The

widening of the life expectancy gap between currently sovereign and affiliated states accelerated in

the 1990’s and continues to increase.

Strengths and limitations

The main strength of this study is that the time period for which life expectancy data were available

encompasses the complete period of Caribbean decolonization in the last half of the 20th century.

This enabled us to evaluate trend changes before, during and after independence.

One limitation of our study is that the World Population Prospects data, which are widely used,

are of uncertain validity. The Appendix includes a description of the methodology used by the UN

to derive life expectancy data for each Caribbean state. Estimates obtained by the direct method

are considered more reliable than those obtained by the indirect method. For the 18 Caribbean

states included in our analysis, life expectancy estimates of 11 states were based on the direct

method (Appendix table A12). This implies that our conclusions are likely to be robust. Another

limitation is that annual life expectancy data were based on interpolation from 5-year estimates.

Since interpolation smooths year-to-year fluctuations, we may have missed less pronounced

changes in annual life expectancy coinciding with decolonization.

GDP per capita data were subject to data availability; the first available data year was 1960. We

therefore calculated economic growth in the pre-independence decade based on three years only

for Jamaica and Trinidad & Tobago and seven years for Barbados and Guyana. Our conclusions

are, however, likely not sensitive to the exact 10-year period changes in these states; our results

show that, relative to the currently affiliated states, their GDP per capita trends were already

unfavorable before decolonization (figure A2). Moreover, we accounted for the observation period

in the likelihood analysis.

Our results do not necessarily imply a causal effect; states that became independent may be

different from states that remained affiliated in other respects than the event of decolonization

Chapter 4

84

alone. A previous study found that identity of (former) colonizer and duration of colonization

were important determinants for contemporary outcomes in GDP per capita and infant mortality,

possibly through the establishment of more (or less) inclusive institutions (25). Ten of eleven

currently sovereign states we assessed are former British colonies, but data of states that remained

affiliated to the United Kingdom (UK) were not available for comparative analysis. Given the

methods we used, we were also unable to include duration of colonization in our analyses. We

consider identity of (former) colonizer and duration of colonization, however, unlikely to be

responsible for contemporary life expectancy disparities between currently sovereign and affiliated

states. First, more recent life expectancies in UK-affiliated states from the PAHO-database

resemble those of affiliated states in our analysis (Pan American Health Organization). Second,

decolonization also coincided with reductions in life expectancy growth in the former Dutch

colony Surinam. Third, Caribbean states in our analysis were at least 250 years under imperial rule,

so any differences in length of colonization are unlikely to affect our results.

It should also be noted that caution of interpretation is needed to generalize our results to other

former colonies. Our results may be specific for small island states, since they share characteristics

that pose particular development challenges, such as their small size, their remoteness and their

narrow resource base (26).

Interpretation

To the best of our knowledge, our study is the first to explore the development of population

health and the potential role of changing economic conditions during decolonization in the

Caribbean region. To discuss possible explanations of our results in more detail, we will

differentiate between a before independence, during decolonization and after independence

period.

Surely, decolonization is unlikely to be the cause of the lower life expectancies and GDP per capita

in currently sovereign states before independence. On the contrary, this finding suggests that

Caribbean states with lower development indicators were more likely to become sovereign, as

already noted by Bertram (27). Empirical studies support the idea that economic factors played an

important role in the widespread decolonization events in the 20th century. Imperial countries were

increasingly willing to give up their colonial ‘assets’ when it became more lucrative to trade than

to occupy (28,29). Other studies emphasize the rise of the international free market economy under

United States hegemony (30,31), even as other factors, such as catalytic effects of previous

decolonization (32), play significant roles. As life expectancy development is known to strongly


85

depend on economic conditions (15), this may explain the lower life expectancies in currently

sovereign states before independence.

External pressures from imperial countries are, however, unlikely to be the main cause of

Caribbean decolonization after WWII. The petition for independence was primarily initiated by

local governments, who also largely set the timing and pace to independence (33,34). Perhaps more

importantly, as McElroy and Mahoney concluded in their study on the reluctance of currently

affiliated populations to petition for independence, affiliated states: “remain unwilling to trade the

visible security, affluence and standard of living of affiliation for the less tangible but more costly

rewards of autonomy” (35).

So, what caused the reductions in life expectancy growth that coincided with decolonization? Our

results suggest that generally worsening economic conditions are unlikely to be the main

determinant. In three states, Antigua & Barbuda, Guyana and Jamaica, declines in life expectancy

growth occurred while GDP per capita more than doubled in the decade after independence.

Caribbean decolonization in the last half of the 20st century was also not accompanied by major

social disruptions, with the exception of Belize. Belize’s independence provoked Guatemala’s

renewal of its long-standing claim to the entire territory

(https://en.wikipedia.org/wiki/History_of_Belize#Decolonization_and_the_border_dispute_wi

th_Guatemala), which has likely contributed to the observed reductions in life expectancy growth

in Belize after independence (Appendix table A2), but such disruptions have not occurred

anywhere else. Furthermore, we show that life expectancy developments during decolonization

did not correspond with similar trend changes in infant mortality. This suggests that quality of

mother and child care was relatively stable during decolonization, although experiences of

individual states varied greatly (Appendix table A5).

Descriptions on how political decisions affected healthcare and (colonial) bureaucracies during

decolonization in the Caribbean states are absent in the scientific literature. Nonetheless,

circumstantial evidence points to the weakening of bureaucracies underlying the uneven life

expectancy developments of currently affiliated and sovereign states. First, weakened

bureaucracies during decolonization gave way to the institutionalization of corruption, defined as

the use of state resources for private gain (19). The corruption of governmental officials and the

police force is involved in the trafficking of drugs in the Caribbean (36). Drug trafficking and

related crimes produced a fast escalation in murder rates in sovereign Caribbean states (37), while

the Caribbean was considered relatively safe in the 1960’s and 1970’s (38).

Chapter 4

86

Political corruption has also been linked to reduced economic output, educational output and

social equalities in sovereign Caribbean states (36). These factors, in combination with higher rates

of crime and gun-related violence, would influence the living conditions after independence and

lead to higher stress and uncertainty experienced among the population. This is consistent with

the rise of emigration between 1965 and 2000 and resulted in net population losses in several

currently sovereign states (39). On average, 70% of the migrants were tertiary educated, as

indicated by the popularly used term ‘brain drain’ (40). This large-scale emigration may have also

directly influenced life expectancy growth of the relatively small populations in currently sovereign

states, considering that higher educated people have higher life expectancies. Increased stress and

uncertainty after decolonization may have also led to higher suicide rates and unhealthy coping

behavior, as previously seen during the political transition of post-communist countries (41,42).

Second, weakened bureaucracies challenge the governments’ capacity to influence areas that

contribute to population health. Contemporary reports indicate that the capacity of sovereign

Caribbean governments to generate health information to evaluate population health and the

quality of healthcare remains challenged (43,44). Moreover, while sovereign states adopted a

resolution to address the non-communicable diseases crisis in 2002 (45), more than a decade later

effective responses were not implemented (46). Finally, where bureaucracies are weak, it is the

poor population that suffers the most (47). Poverty in the Caribbean disproportionally affects

women more than men due to lower earnings, poorer working conditions, increased risks of sexual

and physical violence and the burden of caring for a single-parent family (48,49). This may explain

why our results show that declines in life expectancy growth in currently sovereign states were

more pronounced for females than for males relative to the currently affiliated states (table 2).

Importantly, the explanation of the effect of decolonization on life expectancy in currently

sovereign states through weakened bureaucracies does not imply that currently affiliated states do

not experience high levels of corruption and challenges in their capacity to improve population

health. Rather, continuous political affiliation may create more favourable political conditions for

Caribbean states to implement policies that are, directly or indirectly, beneficial for population

health, for example through the involvement of metropolitan countries’ law enforcement and

justice departments to address drug trafficking and gun-related violence in affiliated states (37).

In the decades after decolonization, our results clearly imply that life expectancies followed

different trajectories across Caribbean states with differing political status, most notably that the

widening of the life expectancy gap between sovereign and affiliated states accelerated in the 1990’s

and continues to increase. These widening disparities are likely related to the Latin American debt


87

crisis and related factors which, as our results suggest, impacted sovereign states more than

affiliated states (figure A2). Weakened bureaucracies may not only have made currently sovereign

states more susceptible to the effects of this crisis on their economies (36), but may also have

aggravated its impact on life expectancy development, for example through increased crime rates

and the discontinuation of programs beneficial to population health.

To date, the ongoing development problems in many former colonies remain mired in

controversy. On the one hand, postcolonial critics debate the Western view of development, and

argue that rich countries’ commitment to improve living conditions in former colonies too often

ignores the dependency of these countries on international markets, and does not fundamentally

address the disparities in political and economic power that were present under colonialism (50–

53). On the other hand, the neoliberal view adopted by many economists attributes the ongoing

problems in the Caribbean to local failures such as inadequate spending on social development

(54), persistence of high income inequalities (49), and corruption (36). A detailed discussion of

both view-points is beyond the scope of this paper, but it is apparent that these beliefs influence

contemporary debates on development in many former colonies.

By no means does our study intend to imply that a return to colonial status is preferred based on

the higher development indicators of currently affiliated Caribbean states. The impact of

decolonization, however, merits analysis for the sake of explaining the long-term effects of the

transition to independence on contemporary health outcomes. Further studies could shed light on

the institutions, policies and other events that contributed to the uneven life expectancy

developments in the Caribbean. Additional research on the counterfactual experience of Cuba

seems particularly interesting, since it is the sole sovereign state that has seen major advances in

population health during the decades after the revolution (55,56), despite having comparable

economic development to other sovereign Caribbean states (57).

Implications and general conclusion

Despite considerable life expectancy gains in all Caribbean states, life expectancy in currently

sovereign states increasingly lags behind that of states which remained affiliated. Our results

indicate that changing economic conditions and violent eruptions were not the main determinant

of the unfavorable trends in life expectancy during and after decolonization. Circumstantial

evidence points to the weakening of bureaucracies during decolonization underlying the uneven

life expectancy developments of currently affiliated and sovereign states.

Chapter 4

88

Since the state is the main unit of political decision-making, continuous political affiliation may

create more favourable political conditions for Caribbean states to implement policies that are,

directly or indirectly, beneficial for population health. Further studies could shed light on the

institutions, policies and other events that contributed to the uneven life expectancy developments

in the Caribbean.

Appendix



89

References

1. Hambleton IR, Howitt C, Jeyaseelan S, Murphy MM, Hennis AJ, Wilks RJ, et al. Trends in Longevity in the Americas: Disparities in Life Expectancy in Women and Men, 1965-2010. PLoS One. 2015;10(6).

2. Muntaner C, Borrell C, Ng E, Chung H, Espelt A, Rodriguez-Sanz M, et al. Politics, welfare regimes, and population health: controversies and evidence. Sociol Health Illn. 2011;33(6):946–64.

3. Mackenbach JP. Political conditions and life expectancy in Europe, 1900-2008. Soc Sci Med. 2013;82:134–46.

4. Beckfield J, Krieger N. Epi + demos + cracy: Linking Political Systems and Priorities to the Magnitude of Health Inequities—Evidence, Gaps, and a Research Agenda. Epidemiol Rev. 2009;31(1):152–77.

5. Navarro V, Shi L. The political context of social inequalities and health. Soc Sci Med. 2001;52(3):481–91.

6. Navarro V, Muntaner C, Borrell C, Benach J, Quiroga A, Rodríguez-Sanz M, et al. Politics and health outcomes. Lancet. 2006 Sep 16;368(9540):1033–7.

7. Pickett KE, Wilkinson RG. Income inequality and health: a causal review. Soc Sci Med. 2015;128:316–26.



10. Iyer L. The long-term impact of colonial rule: Evidence from India. Harvard Bus J. 2004;Mimeo.

11. Dominicus DA, Akamatsu T. Health policy and implementations in Tanzania. Keio J Med. 1989;38(2):192–200.

12. Flanagin A. Letter from Zimbabwe: Health Care Changes After Independence and Transition to Majority Rule. J Am Med Assoc. 1990;263(5):638–40.

13. Sylwester K. Decolonization and Economic Growth: the Case of Africa. J Econ Dev. 2005;30(2):87–102.

14. Rodrik D. Policy uncertainty and private investment in developing countries. J Dev Econ. 1991;36(2):229–42.


16. Mackenbach JP, Looman CW. Life expectancy and national income in Europe, 1900-2008: an update of Preston’s analysis. Int J Epidemiol. 2013;42(4):1100–10.

17. Collier P, Elliot VB, Hegre H, Hoeffler A, Reynald-Querol M, Sambanis N. Breaking the conflict trap. Civil war and development policy. Washington D.C.: World Bank/Oxford University Press; 2003.

18. Ghobarah HA, Huth P, Russett B. The post-war public health effects of civil conflict. Soc Sci Med. 2004;59(4):869–84.

19. Kenny PD. Colonial rule, decolonisation, and corruption in India. Commonw Comp Polit.

Chapter 4

90

2015;53(4):401–27.

20. Warren A. Medicine and Politics in Colonial Peru. 1st ed. University of Pittsburgh Press; 2010.

21. Arnold PC. Why the ex-colonial medical brain drain? J R Soc Med. 2011;104(9):351–4.

22. United Nations: Department of Social and Economic Affairs. World population prospects: The 2012 revision, DVD edition [Internet]. 2013 [cited 2015 May 1]. Available from: http://esa.un.org/unpd/wpp/Excel-Data/population.htm

23. Bulmer-Thomas V. The Economic History of the Caribbean since the Napoleonic Wars. 3rd ed. London: Cambridge University Press; 2014.

24. Hoffmann R, Plug I, Khoshaba B, McKee M, Mackenbach JP. Amenable mortality revisited: The AMIEHS study. Gac Sanit. 2013;27(3):199–206.

25. Feyrer J, Sacerdote B. Colonialism and Modern Income – Islands as Natural Experiments. Cambridge; 2006. (NBER Working Paper Series). Report No.: 12546.

26. United Nations. The Future We Want: Outcome document adopted at Rio+20 Conference on Sustainable Development. In: Rio de Janeiro Meeting. 2012.


28. Galbraith JK. A Journey Through Economic Time: A Firsthand View. Cambridge: Houghton Mifflin Harcourt; 1995.

29. Holland R. European decolonization, 1918-1981: an introductory survey. 1985th ed. London: Palgrave; 1985.

30. Gartzke E, Rohner D. The political economy of imperialism, decolonization and development. Br J Polit Sci. 2011;41(03):525–56.

31. Grossman HI, Iyigun MF. Population Increase and the End of Colonialism. Economica. 1997;64(255):483–93.

32. Strang D. Adding social structure to diffusion models: an event history framework. Sociol Methods Res. 1991;19(3):324–53.



35. McElroy JL, Mahoney M. The propensity for political dependence in island microstates. Insul Int J Isl Aff. 1999;9:32–5.

36. Collier MW. The effects of political corruption on Caribbean development. Miami: Proceedings of the Caribbean Studies Annual Conference; 2002.


38. de Albuquerque K. A comparative analysis of violent crime in the Caribbean. Soc Econ Stud. 1984;33(3):93–142.

39. St. Bernard G. Major trends affecting families in Central America and the Caribbean. St. Augustine,


91

Trinidad & Tobago: United Nations Division of Social Policy and Development; 2003.

40. Mishra P. Emigration and brain drain: evidence from the Caribbean. 2006. (IMF Working Paper). Report No.: WP/06/25.

41. Varnik A, Kolves K, Sisask M, Samm A, Wasserman D. Suicide mortality and political transition: Russians in estonia compared to the Estonians in Estonia and to the population of Russia. Trames. 2006;10(3):268–77.

42. Leon D, Chenet L, Shkolnikov V, Zakharov S, Shapiro J, Rakhmanova G, et al. Huge variation in Russian mortality rates 1984-94: artefact, alcohol, or what? Lancet. 1997;350(9075):383–8.

43. Eldemire-Shearer D. Health information systems in the Caribbean: time for attention. West Indian Med J. 2008;57(4):321–2.

44. Cunningham-Myrie C, Reid M, Forrester T. A comparative study of the quality and availability of health information used to facilitate cost burden analysis of diabetes and hypertension in the Caribbean. West Indian Med J. 2008;57(4):383–92.

45. Pan American Health Organization (PAHO). Public Health Response to Chronic Diseases. In: PAHO/WHO, editor. 26th Pan American Sanitary Conference, 54th session of the Regional Committee. Washington D.C., USA: PAHO/WHO; 2002.


47. Glaeser E, Scheinkman J, Shleifer A. The injustice of inequality. J Monet Econ. 2003;50(1):199–222.

48. Christian Aid. The Scandal of Inequality in Latin America and the Caribbean [Internet]. 2012. Available from: https://www.christianaid.org.uk/images/scandal-of-inequality-in-latin-america-and-the-caribbean.pdf

49. de Ferranti D, Perry GE, Ferreira FHG, Walton M, Coady D, Cunningham W, et al. Inequality in Latin America and the Caribbean: breaking with History. Washington D.C.: World Bank; 2003.

50. Cardoso FH, Faletto E. Dependency and Development in Latin America. 1st ed. University of California Press; 1979.

51. Nkrumah K. Neo-Colonialism, The Last Stage of Imperialism. New York: International Publishers; 1966.

52. Dirks NB. Colonial and Postcolonial Histories: comparative reflections on the legacies of empire. 2004. (Global Background Paper for the Human Development Report 2004).

53. Chiriyankandath J. Colonialism and Post-Colonial Development. In: Burnell P, Randall V, Rakner L, editors. Politics in the Developing World. Third. New York: Oxford University Press; 2011.

54. OECS. Human Development Report 2002 Building competitiveness in the face of vulnerability. 2002.

55. Cooper RS, Kennelly JF, Ordunez-Garcia P. Health in Cuba. Int J Epidemiol. 2006;35(4):817–24.

56. Ebanks GE. Contrasting Population Challenges: Cuba and the Dominican Republic in the Next Millenium. 1998. (PSC Discussion Papers Series: Vol. 12: Iss.9, Article 1; vol. 12).

57. Mesa-Lago C. Growing economic and social disparities in Cuba: impact and recommendations for change. Pittsburgh: Institute for Cuban and Cuban-American Studies, University of Miami; 2002.

Chapter 5

Differences in life expectancy

between four Western countries and their Caribbean

dependencies,

1980-2014


European Journal of Public Health 2019, pii: ckz102

https://doi: 10.1093/eurpub/ckz102

Chapter 5

94

Abstract

Background In the Caribbean, life expectancy in politically independent territories has

increasingly diverged from that of territories that remained affiliated to their former colonizers.

Because these affiliated territories differ in degree of political independence, they are not all

governed in the same way. We assessed whether differences in life expectancy trends between

Caribbean dependencies and their Western administrators were related to their degree of

political independence, and which causes of death contributed to divergence or convergence in

life expectancy.

Methods Analysis of age-standardized death rates and decomposition of life expectancy

differences between France, the Netherlands, the UK, the USA and their Caribbean

dependencies by age and cause-of-death during the period 1980-2014.

Results Life expectancy differences between Western countries and their dependencies have

generally increased for men and narrowed for women, but trends have been much more

favorable in the French- than in the Dutch-administered territories. The strongest contributions

to widening gaps in life expectancy between Western countries and their dependencies were

from mortality from cardiovascular diseases (ischemic heart disease) and external causes

(homicide and traffic accidents).

Conclusion Dependencies with a stronger political affiliation to a Western country

experienced more favorable life expectancy developments than dependencies that had more

autonomy during the 1980-2014 period. The underlying mortality differences with Western

countries are largely comparable among Caribbean territories but differ in magnitude, most

notably for cardiovascular disease and external causes. This suggests that increases in a

territory’s political autonomy impairs the diffusion of new knowledge and techniques, and/or

reduces government’s effectiveness in implementing policies.

Differences in life expectancy Western countries and Caribbean dependencies, 1980-2014

95

Introduction

Since the 1960s, when many Caribbean island territories gained independence, life expectancy

trends in the Caribbean have diverged. As a consequence, the Caribbean region currently has the

lowest life expectancies and largest between-country disparities in life expectancy of the Americas

(1). Previous studies observed large differences in life expectancy development between dependent

and independent Caribbean territories (2–4), which may be the result of two self-reinforcing

mechanisms. On the one hand, the weakening of bureaucracies after decolonization challenge a

governments' capacity to influence areas that contribute to population health (5). On the other

hand, affiliation to some of the most advanced countries in the world (France, the Netherlands,

the United Kingdom (UK) and the United States of America (USA)) likely promotes a process of

policy diffusion that is, directly or indirectly, beneficial for population health (6). In other words,

an absence of the disadvantages of sovereignty, and the presence of the advantages of a strong

political, cultural, and (socio)economic relationship with a prosperous country, probably results in

the relatively high life expectancies in affiliated territories. We will refer to “affiliated territories”

and “dependencies” as neutral terms to describe the political status of Caribbean territories without

sovereignty under international law, as indicated by their non-membership of the United Nations

(http://www.un.org/en/member-states/index.html).

Life expectancies, however, vary greatly between Caribbean dependencies (7). Possibly underlying

these differences is a gradient in autonomy caused by the way Western countries govern their

dependencies, which varies considerably and also translates into a variety of governance structures

on the area of public health. At one end of the autonomy-dependency spectrum, we find

Guadeloupe and Martinique. These French departments are considered an integral part of the

French Republic. Health system governance on the islands is implemented with oversight from

the Ministry of Health in metropolitan France, since 2010 by the ‘Agence Régionale de Santé’

(ARS). Next are the small island territories of the UK: Anguilla, British Virgin Islands, Cayman

Islands, Montserrat and Turks & Caicos. All islands have their own constitution, legal system and

a democratically elected Government (with the exception of Turks & Caicos during the 2009-2012

period). While the island governments are considered to have primary responsibility for the

provision of healthcare and public health interventions to their populations, the UK government

leads with overall policy through the Department of Health, the Department for International

Development and the Health Protection Agency. They pro-actively support the islands to manage

their health sector, to work with regional organizations, to prepare for emergencies and to fulfil

international responsibilities such as the International Health Regulations (8). Succeeding these

Chapter 5

96

islands are Puerto Rico and the US Virgin Islands, two unincorporated territories of the USA.

Whereas Puerto Rico is a self-governing territory based on its commonwealth status, the US Virgin

Islands is considered not. Health governance, however, is similarly organized and provided by the

local Departments of Health that lead all efforts related to health on the islands. At the other end

of the spectrum, we find the constituent countries of the Kingdom of the Netherlands: Aruba, the

Netherlands Antilles (which consisted of five island territories and was dissolved in 2010) and

Curaçao (the most populous island territory of the former Netherlands Antilles). Similar to the

UK islands, the Dutch territories have their own constitutions, legal systems and a democratically

elected Government. In contrast to the UK islands, however, the Dutch islands govern their own

health sectors and steer their own health policy initiatives.

In this paper, we will assess whether Caribbean dependencies with a greater degree of political

independence have less favourable life expectancy trends, as measured by the development over

time of the gap in life expectancy with their Western administrators. In order to obtain insight in

the factors underlying convergence or divergence of life expectancy we will also determine the

contribution of specific age-groups and causes of death to gaps in life expectancy.

Methods

Data

We selected all Caribbean dependencies for which mortality and population data were available

during the 1980-2014 period. This excluded one territory of France and four territories of the

Netherlands. We obtained mortality data from the World Health Organization (WHO) mortality

database (9) or through a request at the Pan American Health Organization (PAHO) for Curaçao.

Data include annual deaths by gender and age in each year, using the ninth or tenth revisions of

the International Classification of Diseases (ICD). ICD-codes are given in Appendix Table A1.

Demographic data by gender and age were extracted from the United Nations (UN) World

Population Prospects (WPP) 2017(10) or, for islands with 2015 populations smaller than 60 000,

from the US Census Bureau (11). In preliminary analyses we noticed that records concerning

certain causes of death were missing in the datasets for Aruba prior to 1995. This involved the E-

codes indicating external causes, for which most sub-causes registered an implausible count of

zero. We continued analyses as normal and took these omissions into account during the

interpretation of the results.


97

Data analyses

We totalled mortality and population data over 5-year periods to reduce sensitivity to small number

fluctuations. For the same reason, we combined all deaths in the five dependencies of the UK.

The actual number of data years included per 5-year period varies due to data availability. We

corrected deaths for unclassified age and gender, for deaths attributed to ill-defined causes (R00-

R99, non-external deaths only) and for under-registration (as determined by comparing the

number of deaths with independent estimates from the WPP 2017 and the US Census Bureau)

(12). Based on the corrected numbers of mortality, we calculated age-standardized mortality rates

per 100 000 population using direct standardization to the World standard population (13). We

computed life expectancy from abridged life tables and the age- and cause-specific contributions

to life expectancy differences by Arriaga’s decomposition method (14) and calculated 95%

confidence limits using Monte Carlo simulations (15).

Results

As figure 1 shows, life expectancy in France, the UK, the USA and the Netherlands has increased

steadily since the early 1980s. Life expectancy in their Caribbean dependencies has also increased,

but in a less regular pattern. In all Caribbean dependencies, the relative position in male life

expectancy with respect to their Western administrators deteriorated. In contrast, the gap in female

life expectancy that existed in the early 1980s has improved in all dependencies but Aruba

(Appendix table A2). In the Caribbean dependencies of France and the UK, life expectancy has

moved more or less in parallel with that in their administrating countries, and as a result gaps in

life expectancy were relatively small or non-existent in the early 2010s. For the UK islands this

meant that they lost their initial advantage in male life expectancy relative to the UK. Trends in life

expectancy were less favorable in the Caribbean dependencies of the Netherlands and the USA.

In Aruba and the Netherlands Antilles, male life expectancy decreased and female life expectancy

stagnated during the 1990s, which resulted in a substantial gap in life expectancy with the

Netherlands in the early 2010s. A smaller gap in male life expectancy has opened up between

Puerto Rico and the USA. Similar to the trends in the Dutch islands, however, male life expectancy

decreased in the US Virgin Islands which resulted in a gap of almost 5 years with the USA in the

early 2010s. Remarkably, due to faster increases over the study-period, female life expectancy in

Puerto Rico and US Virgin Islands is now higher or comparable, respectively, than that in the

USA.

Chapter 5

98

Figure 1 Life expectancy at birth in Western countries and their Caribbean dependencies, 1980-2014


99

The differences in male life expectancy between Caribbean dependencies and their administrating

countries have been decomposed by age and cause-of-death after combining the populations and

deaths of the dependencies of each Western country, both for the earliest and the latest available

time period. In the top of figure 2, we again see that life expectancy differences have generally

increased for men: the difference between the top and bottom column is always negative. The

magnitude of this increase differs greatly between Caribbean dependencies (France: 1.0 years, UK:

2.8 years, US: 2.2 years, the Netherlands: 4.4 years). When considering age-specific contributions

to life expectancy differences, we observe similarities in the direction of the changes during the

study period. In all Caribbean dependencies, infant mortality trends have been more favorable

than in their Western administrator. For young and elderly men, the opposite is true and the gap

between Western countries and dependencies has become larger. In all dependencies except the

French, mortality trends among middle-aged men have been unfavorable compared to their

Western administrators.

When considering cause-specific contributions to male life expectancy differences, we see that the

strongest contributions are from diverging trends in cardiovascular diseases and external causes

(Appendix table A3). The cardiovascular mortality gap has increased in all Caribbean dependencies

except the French, with the greatest increases seen in the Dutch territories. Within this category

this is mainly caused by ischemic heart diseases, whose initial advantage in the dependencies almost

completely disappears during the study period. Next in magnitude are changes in external causes,

whose contribution to life expectancy differences with Western countries increased in all

dependencies, but again most strongly in the Dutch territories. The leading causes in this category

are transport accidents and homicide. Additional causes that had a smaller contribution to the

increasing gap in male life expectancy are neoplasms (prostate and lung cancer), diabetes and

infections (HIV/AIDS).

Cha

pter

5

100

Fig

ure

2Co

ntrib

utio

n to

life

exp

ecta

ncy

diffe

renc

e be

twee

n W

este

rn c

ount

ries a

nd th

eir C

arib

bean

dep

ende

ncies

, ear

liest

vs.

lates

t ava

ilabl

e tim

e pe

riod,

men

Diff

eren

ces

in li

fe e

xpec

tanc

y W

este

rn c

ount

ries

and

Car

ibbe

an d

epen

denc

ies,

198

0-20

14

101

F

igu

re 2

(Con

tinue

d)

Chapter 5

102

Differences in female life expectancy between Western countries and their dependencies decreased

over the study period (Appendix figure A1), but the magnitude of these gains are relatively modest

(France: -1.2 years, UK: -3.0 years, US: -1.5 years, the Netherlands: -1.6 years). Relative to

developments in Western administrators, trends in cardiovascular mortality have been more

favorable in all dependencies except the Dutch. For external mortality, the gap between Western

countries and their dependencies remained similar to the beginning of the study period, but

increased in the Dutch territories. Additionally, the impact of mortality from diabetes increased in

all dependencies except the French islands. These contributions are, however, small relative to

those that occurred in favor of female life expectancy convergence between Caribbean

dependencies and Western administrators.

Finally, we look at trends in age-standardized mortality from cardiovascular diseases and external

causes. As figure 3 shows, male mortality from cardiovascular diseases decreased steadily in

Western countries since the early 1980s. In the dependencies of France, the UK and the USA,

cardiovascular mortality trends largely followed those of their Western administrator. In stark

contrast, cardiovascular mortality has poorly improved in the Dutch dependencies. For women,

trends in cardiovascular mortality are similar (Appendix figure A2). Male mortality trends from

external causes show large differences between the Western countries and their dependencies

(figure 3). In the Netherlands Antilles/Curaçao, Puerto Rico, and the US Virgin Islands, this was

due to dramatic increases in homicide mortality during the study period. Transport accidents are

the main contributor to external mortality in Guadeloupe, Martinique, Aruba and the UK islands,

and an important additional contributor to external mortality in the Netherlands Antilles/Curaçao

(Appendix table A4). Differences in female mortality from external causes between Western

countries and their dependencies are relatively small (Appendix figure A2).

Discussion

Limitations

This analysis is subject to several limitations. The most obvious relates to the availability and quality

of mortality data in Caribbean territories. Ideally, all data years would be available for all

dependencies, but this was not the case. To make best use of the available data, we pooled the data

in 5-year periods, but may consequently have missed less pronounced trend changes. We corrected

registered deaths for data inadequacies and incompleteness using PAHO methods (12). Compared

to Western countries, mortality data of Caribbean dependencies had more inadequacies and were


103

Figure 3 Trends in age-standardized mortality rates from cardiovascular diseases and external causes in Western countries and their Caribbean dependencies, 1980-2014, men

Chapter 5

104

more incomplete, except in Puerto Rico and the US Virgin Islands (Appendix table A5). With the

PAHO corrections we have minimized the effects of data quality issues as much as possible.

The corrections, however, assumed that misclassified and missing deaths from the vital statistics

followed the same distribution as recorded deaths and did not differ by age, gender and cause-of-

death. These assumptions were, however, not met in the mortality registration of Aruba prior to

1995, in which several external causes were omitted. Although this has led to an underestimation

of the total mortality and external causes in particular, we judge it unlikely that our conclusions are

mainly attributable to this registration bias, as the results during this period are generally consistent

with the years after 1995.

Another limitation is that many Caribbean territories have experienced population losses caused

by large-scale emigration (16). This may plausibly have effected mortality patterns, because

migrants are often well-educated and young (17). The assumption that population declines have

influenced life expectancy trends was tested in an additional analysis (Appendix table A6). We

found that the largest fluctuations in population size occurred in Aruba, Curaçao and the UK

islands. In Curaçao the population mainly declined in the 1980-2014 period, while in Aruba and

the UK islands it mainly increased. Since unfavorable life expectancy trends were found in both

Aruba and Curaçao, we have no indication that population declines have influenced life expectancy

developments in these territories.

Interpretation

Our findings suggest that the way that Western countries govern their dependencies has strongly

influenced life expectancy trends during the 1980-2014 period. Life expectancy developments in

Caribbean dependencies that had relatively low levels of autonomy, Martinique, Guadeloupe and

the UK islands, ran relatively in parallel with that in their administrator countries. In contrast, the

US territories with more autonomy, Puerto Rico and the US Virgin Islands, experienced declines

in male, and developments more similar to the US in female life expectancy. The constituent

countries of the Kingdom of the Netherlands, Aruba and the Netherlands Antilles/Curaçao,

experienced declines in male and stagnation in female life expectancy. This shows that territories

with a stronger political affiliation experienced more favourable life expectancy developments.

Our conceptual model focuses on two macro-level explanations of health disparities between

affiliated and more autonomous Caribbean states: stronger political affiliation increases a

government’s capacity to influence determinants of population health (i.e., the ability to implement


105

and enforce policies) and facilitates diffusion of the knowledge and tools necessary to implement

effective policies. We see these ‘upstream’ factors as defining the context in which ‘downstream’

determinants of population health develop, such as medical practice, prevention campaigns, and

the health behaviours of individuals, which more directly impact risks of mortality.

Decomposition of male life expectancy differences between Western countries and their

dependencies point to the important contribution of mortality from cardiovascular diseases and

external causes. The underlying patterns of trend changes relative to their Western administrators

are similar among Caribbean states and possibly reflect common region-specific drivers, such as a

high prevalence of obesity and other non-communicable disease risk factors (18) and a strong

presence of the international drug trade and weapon trafficking (19) that involves corrupt

governmental officials (20). We also find, however, that Caribbean dependencies differ in the

magnitude of trend changes, again with more favourable trends in the French than in the Dutch

administered territories.

Unlike other islands, Dutch dependencies did not experience substantial improvements in

cardiovascular mortality, which is in line with the high levels of uncontrolled hypertension that

were measured during a recent health survey in Curaçao (21). This suggests that the third

epidemiological transition, characterized by progressive declines in cardiovascular mortality (22),

is delayed in the Dutch dependencies. In the Netherlands, cardiovascular mortality has gradually

declined since its peak in the late 1950s for women and early 1970s for men (23). This suggests

that the Dutch Caribbean populations have not fully taken advantage of advancements in

prevention and treatment of cardiovascular diseases and their underlying causes.

The contribution of external causes to life expectancy differences between Western countries and

dependencies has also increased, most strongly in Dutch (both genders) and US (men only)

territories. In the Netherlands Antilles/Curaçao, Puerto Rico, and US Virgin Islands, this

divergence was due to sharp rises in homicide, which gave these territories the dubious distinction

of being among those with the highest homicide rates in the world (24). Transport accidents are

the main contributor to external cause mortality in Aruba and an important additional contributor

to external cause mortality in the Netherlands Antilles/Curaçao. Just like after decolonization

(5,25), increases in a dependency’s autonomy may weaken bureaucracies and subsequently

challenge its government’s capacity to address organized crime and to implement and enforce

traffic regulations.

Chapter 5

106

The less favourable trends for men than for women suggest that men are more sensitive to the

negative effects of political sovereignty. While we have no specific data to shed light on this

intriguing finding, we note that the life expectancy declines in former USSR countries after the

economic and political shocks caused by the fall of communism were also largely limited to men

(26). Somehow, men seem to respond differently to unfavourable social or economic conditions

than women, with a greater tendency to engage in potentially lethal behaviour. In the case of the

Caribbean, economic shocks are likely to be of less importance for the explanation of our findings.

While life expectancy trends have been least favorable in the Dutch and US territories, we did not

find an indication of reduced economic growth in these territories relative to those of France and

the UK (Appendix table A7).

As far as we are aware, our study is the first attempt to explore the role of political conditions on

life expectancy developments in Caribbean dependencies. We recognize that the current political

status of these dependencies is in many aspects comparable to the colonial era, and could in fact

be seen as a continuance of the asymmetrical power relationship with a former colonizer, as is

regularly echoed in political rhetoric (27,28). Contemporary literature points to the benefits of late

colonialism largely outweighing its disadvantages (29–32), while also demonstrating that Caribbean

people in Western countries experience negative treatment based on their skin color (33,34), and

are disproportionally disadvantaged for factors that contribute to their health (35,36). Considering

the duality of these cultural dynamics, with sustainable development on one side, and social

stigmatization on the other, it may then not be too surprising that the populations of many former

colonies are hesitant to allow more “Western” influences in their internal affairs, yet currently

struggle to move forward as well.

Appendix



107

References

1. Hambleton IR, Howitt C, Jeyaseelan S, Murphy MM, Hennis AJ, Wilks RJ, et al. Trends in Longevity in the Americas: Disparities in Life Expectancy in Women and Men, 1965-2010. PLoS One. 2015;10(6).




5. Verstraeten SPA, van Oers HAM, Mackenbach JP. Decolonization and life expectancy in the Caribbean. Soc Sci Med. 2016;170:87–96.

6. Vallin J, Meslé F. Convergences and divergences in mortality. A new approach to health transition. Demogr Res. 2004;10(Suppl.2):11–44.

7. Pan American Health Organization. PLISA Health Information Platform for the Americas [Internet]. 2019 [cited 2019 Aug 4]. Available from: http://www.paho.org/data/index.php/en/indicators/visualization.html

8. Department for International Development. Health and healthcare in the British Overseas Territories: Regional and UK government support. 2010.



11. US Census Bureau. International database [Internet]. 2018 [cited 2018 Apr 19]. Available from: http://www.census.gov/population/international/data/idb/metadata.php?R=World&C=Anguilla

12. Silvi J. On the estimation of mortality rates for countries of the Americas. Epidemiol Bull Pan Am Heal Organ. 2003;24(4):1–5.

13. Ahmad OB, Boschi-Pinto C, Lopez AD, Murray CJL, Lozano R, Mie I. Age Standardization of Rates: A New WHO Standard. 2001. (GPE Discussion Paper Series). Report No.: 31.

14. Auger N, Feuillet P, Martel S, Lo E, Barry AD, Harper S. Mortality inequality in populations with equal life expectancy: Arriaga’s decomposition method in SAS, Stata, and Excel. Ann Epidemiol. 2014;24(8):575–80.

15. Andreev EM, Shkolnikov VM. Spreadsheet for calculation of confidence limits for any life table or healthy-life table quantity. MPIDR Technical Report TR-2010-005. Rostock: Max Planck Institute for Demographic Research; 2010.

16. St. Bernard G. Major trends affecting families in Central America and the Caribbean. St. Augustine, Trinidad & Tobago: United Nations Division of Social Policy and Development; 2003.

17. Mishra P. Emigration and brain drain: evidence from the Caribbean. 2006. (IMF Working Paper). Report No.: WP/06/25.

18. Pan American Health Organization. Strategic Plan of Action for the Prevention and Control of Non-Communicable Diseases for Countries of the Caribbean Community (CARICOM) 2011-2015.

Chapter 5

108

2011;


20. Collier MW. The effects of political corruption on Caribbean development. Miami: Proceedings of the Caribbean Studies Annual Conference; 2002.

21. Verstraeten SPA, Jansen I, Pin R, Brouwer W. De Nationale Gezondheidsenquête 2013: methodologie en belangrijkste resultaten. Willemstad, Curaçao: Volksgezondheid Instituut Curaçao; 2013.

22. Barrett R, Kuzawa CW, McDade T, Armelagos GJ. Emerging and Re-emerging Infectious Diseases: The Third Epidemiologic Transition. Annu Rev Anthropol. 1998;27:247–71.

23. Leening MJG, Siregar S, Vaartjes I, Bots ML, Versteegh MIM, van Geuns RJM, et al. Heart disease in the Netherlands: A quantitative update. Netherlands Hear J. 2014;22(1):3–10.

24. United Nations Office on Drugs and Crime. Global study on homicide. Trends, contexts, data. Vienna: UNODC; 2011.

25. Kenny PD. Colonial rule, decolonisation, and corruption in India. Commonw Comp Polit. 2015;53(4):401–27.

26. Brainerd E, Cutler DM. Autopsy on an Empire: Understanding Mortality in Russia and the Former Soviet Union. 2004. (NBER Working Paper). Report No.: 10868.

27. Cardoso FH, Faletto E. Dependency and Development in Latin America. 1st ed. Berkeley: University of California Press; 1979.


29. De Juan A, Pierskalla JH. The comparative politics of colonialism and its legacies: an introduction. Polit Soc. 2017;45(2):159–72.

30. Woodberry RD. The missionary roots of liberal democracy. Am Polit Sci Rev. 2012;106(2):244–74.

31. Acemoglu D, Johnson S, Robinson JA. The colonial origins of comparative development: an empirical investigation. American Economic Review. Cambridge; 2000. (NBER; vol. 91). Report No.: 7771.


33. Andriessen I, Fernee H, Wittebrood K. Ervaren discriminatie in Nederland. Den Haag: Sociaal en Cultureel Planbureau; 2014.

34. Taylor RJ, Forsythe-Brown I, Mouzon DM, Keith VM, Chae DH, Chatters LM. Prevalence and correlates of everyday discrimination among black Caribbeans in the United States: the impact of nativity and country of origin. Ethn Heal. 2019;24(5):463–83.

35. Verstraeten SPA, van den Brink CL, Mackenbach JP, van Oers HAM. The health of Antillean migrants in the Netherlands: a comparison with the health of non-migrants in both the countries of origin and destination. Int Health. 2018;10(4):258–67.

36. Hickling FW, Rodgers-Johnson P. The incidence of first contact schizophrenia in Jamaica. Br J Psychiatry. 1995;167(2):193–6.

Chapter 6

Health policy performance in 16 Caribbean states,

2010-2015


American Journal of Public Health 2019, vol. 109, pp. 626-632

https://doi.org/10.2105/AJPH.2018.304733

Chapter 6

112

Abstract

Objectives To determine whether Caribbean states vary in health policy performance in 11

different areas; to explore the association with sociodemographic, economical, and governance

determinants; and to estimate the potential health gains of “best-practice” health policies.

Methods We selected 50 indicators that included data on mortality (latest available, 2010–

2015), intermediate outcomes, and policy implementation to calculate a state’s health policy

performance score. We related this score to country characteristics and calculated the potential

number of avoidable deaths if the age-specific mortality rates of best-performer Martinique

applied in all states.

Results We found large differences in health policy performance among Caribbean states.

Martinique, Cuba, and Guadeloupe had the highest performance scores, and Guyana, Belize,

and Suriname the lowest. Political affiliation, religious fractionalization, corruption, national

income, and population density were associated with health policy performance. If the mortality

rates of Martinique applied to all Caribbean states, an overall mortality reduction of 12% would

be achieved.

Conclusions Differences in health outcomes between Caribbean states are partly attributable

to variations in health policy implementation. Our results suggest that many deaths can be

prevented if Caribbean governments adopt best-practice policies.

Health policy performance in 16 Caribbean states, 2010-2015

113

Introduction

Substantial evidence shows that the implementation of effective health policies leads to declines

in mortality from causes that are amenable to policy interventions. For example, implementation

of alcohol control policies has effectively reduced alcohol-related mortality in many populations

(1). Maternal and neonatal mortality rates have declined through policies aiming at safer pregnancy

and childbirth (2). Cervical and breast cancer mortality rates declined after the introduction of

cancer screening programs (3,4).

In the Caribbean, however, a decade after local governments acknowledged the severity of the

region’s health crisis and committed to address their populations’ needs, effective health policies

were still not implemented (5). Caribbean states also vary greatly in population health outcomes,

for example in life expectancy (6), cervical cancer morbidity and mortality (7), adolescent health

(8), and homicide rates (9). In this article, we combined data on health policy implementation with

data on health outcomes to create a comprehensive picture of health policy performance in all

Caribbean states with available data.

Our purpose was to inform policy strategies aimed at improving population health in the

Caribbean by carrying out a broad cross-country comparison of health policy performance and by

identifying Caribbean states that perform better or worse than others. Our study covers 11 policy

areas related to the regional health objectives of the Pan American Health Organization (PAHO):

1. “HIV/AIDS”, 2. “Communicable disease”, 3. “Cancer screening”, 4. “Tobacco”, 5. “Fertility,

pregnancy and childbirth”, 6. “Child health”, 7. “Diabetes”, 8. “Hypertension”, 9. “Alcohol”, 10.

“Road safety”, and 11. “Violence”. We also explored the demographic, economical, health system,

political, governance and cultural determinants of variations in health policy performance within

the Caribbean region. Finally, we calculated the potential health gains if all Caribbean states would

implement best-practice health policies.

Materials and methods

Data

We defined the Caribbean by PAHO's definition and included Belize because of its strong

historical and economic ties to the Caribbean. The source of mortality data was the World Health

Organization’s (WHO) mortality database (10). A detailed description of the calculation of the

mortality rates is given in the Appendix. In short, considering that many Caribbean states have

small populations, we included up to 5 years of the most recent data during the 2010-2015 period

Chapter 6

114

for each Caribbean state. We then systematically assessed whether the mortality data of each state

was suitable for inclusion in our analysis, using three criteria: 1. the total number of deaths was

500 or more, 2. causes of death were registered for at least 70% of deaths in the vital registration,

and 3. quality assessment of the cause-of-death data according to WHO’s ANACoD program was

found sufficient (11). We judged mortality data of 16 Caribbean states suitable for inclusion. Data

sources for policy implementation and intermediate outcome indicators are listed in the notes of

table A3 (Appendix). Data sources for national determinants are listed in the notes of table 1.

Methods

We followed the methods of Mackenbach and McKee, which are briefly described below (12,13).

For each of the 16 Caribbean states, we searched data on policy implementation and intermediate

outcome indicators. We aimed to include at least one indicator on policy implementation, one

indicator on intermediate health outcomes and one indicator on final health outcomes for each

policy area. Policy implementation indicators measure the degree to which a policy is successfully

implemented, e.g. the “alcohol policy sum score” depicts the extensiveness with which

recommendations on effective alcohol policies are implemented. Intermediate outcome indicators

measure health risks that are directly influenced by policy implementation, e.g. total alcohol

consumption among drinkers. Final outcome indicators measure the deaths that could have been

avoided by public health interventions, e.g. mortality from alcohol-related causes. Preferably,

measurements of indicator(s) on policy implementation and intermediate outcomes preceded

measurement of indicator(s) on final outcomes with at least 5 years. Due to limited data availability,

this was not possible for all policy areas (Appendix table A3). We included up to 50 indicators per

country, and were able to cover all 11 policy areas for each country.

As a basis for the calculation of our health policy performance score, we first evaluated whether

policy implementation had likely impacted the level of related health outcomes using correlation

analysis. Next, we calculated a performance summary score per state by assessing whether a state

fell in the lower, intermediate or upper tertile of the Caribbean distribution of each indicator,

followed by subtracting, for each country, the percentage of scores in the upper tertile from the

percentage of scores in the lower tertile. For example, of the 48 indicators that were included for

Cuba, 33 indicators (68.8%) fell in the upper tertile and 8 indicators (16.7%) fell in the lower tertile

of the Caribbean distribution. Cuba’s summary score was therefore 68.8%-16.7%=52.1%

(Appendix table A4). To explore determinants of health policy performance, we related the

summary score to country characteristics using correlation analysis. We attempted to only include


115

data that preceded the policy indicators for at least 5 years, but this was not always possible due to

limited data availability (Table 1).

Finally, for each of the selected causes of death, we calculated the annual average number of deaths

and the Potential Years of Life Lost (PYLL). The PYLL was calculated by taking the difference

between the age of death and age 75. This is a conservative estimate; life expectancy at birth

exceeds 75 years in many Caribbean states. We used the age-specific mortality rates of the state

with the highest summary score, Martinique, to estimate potential health gains per policy area by

calculating the number of deaths that would have occurred if the age-specific mortality rates of

Martinique would have applied, and by subtracting that number from the observed number of

deaths.

Results

Calculation health policy performance summary score

For all health indicators, considerable differences were found between Caribbean states (Appendix

Table A3). Correlations between indicators for policy implementation, intermediate and final

health outcomes were largely in accordance with what one would expect if policy implementation

induces changes in intermediate outcomes, and if changes in intermediate outcomes induce

changes in final outcomes (Appendix Figure A1). For example, HIV/AIDS mortality varied from

1.6 to 45.2 deaths per 100,000 population in a pattern that was consistent with coverage of

antiretroviral therapy (figure A1a). A high score for the International Health Regulations’ core

capacity “Surveillance” was negatively associated with communicable disease mortality in children

(figure A1b) and was also weakly but negatively associated with communicable disease mortality

in adults). This suggests that higher communicable disease mortality could indeed be a

consequence of a state’s substandard ability to monitor infectious disease. The two Caribbean

countries with a cancer policy sum score of 0 because they did not have national policy and action

plan in place and did not provide free cancer-screening programs, Belize and St. Vincent & the

Grenadines, also have relatively high cervix cancer mortality (figure A1c). In Cuba and

Guadeloupe (nearly) all deliveries are attended by trained personnel, and the perinatal mortality

rates (figure A1d) are accordingly low. The indicator “deliveries attended by trained personnel”

also weakly correlates with low maternal mortality and, while not directly related, low levels of

teenage pregnancies. Alcohol consumption is relatively low in Cuba and the Bahamas, which are

both countries with a high number of policy interventions to limit alcohol use (figure A1e). The

Chapter 6

116

more effective a state was in the enforcement of speed laws, the lower the rates of male road traffic

deaths (figure A1f), although this association was weak.

For two policy areas, “tobacco” and “violence”, the policy implementation indicators were not

related to the intermediate and final outcomes in the expected way, but in these cases, based on

previous studies, we nevertheless think that variation in intermediate and final outcomes do reflect

differences in policy implementation (12,14). The number of implemented smoking policies was

weakly but positively associated with the smoking prevalence rate, and the latter was not associated

with lung cancer mortality. This probably indicates that states were more likely to implement

tobacco control measures when smoking prevalence was already high, and that current lung cancer

mortality rates reflect smoking prevalence several decades ago. More restrictions in laws and other

regulations that civilians face to obtain or own a gun are weakly but negatively associated with a

lower number of firearms per 100 population in a territory. However, more restrictive gun policy

is not associated with lower male or female homicide mortality, possibly because of differences in

the enforcement of civilian gun laws and regulations that are not picked up by this policy

implementation indicator.

For three policy areas, “diabetes”, “hypertension” and “child health”, policy implementation

indicators were not available. We therefore had to assume that the outcome indicators indeed

reflect policy implementation, which is not unreasonable in view of the fact that: 1. diabetes and

hypertension screening methods and treatments, as well as dietary salt reduction programs to

prevent hypertension, are applied on a large but varying scale in the Caribbean, and 2. child health

services are typically distributed through governmental programs. A high prevalence of obesity

was associated with raised blood glucose in men but not in women. Raised blood glucose, in turn,

was associated with male and female diabetes mortality. Female stroke mortality was higher in

Caribbean states where the average systolic blood pressure was also high. Measles vaccination

coverage was lower than 90% in Guadeloupe, Jamaica, and Suriname, suggesting that herd

immunity was compromised.

Supported by the associations found between indicators of policy implementation, intermediate

outcomes and final outcomes, we calculated, for each Caribbean state, a health policy performance

summary score that indicates the relative performance of a country for all policy areas together

(Appendix Table A4). In descending order, Martinique, Cuba and Guadeloupe have the highest

summary scores, and Guyana, Belize and Suriname the lowest (Figure 1).


117

Associations health policy performance and country characteristics

In univariate regression analyses we then explored which country characteristics are associated

with variation in summary scores (table 1). We observed statistically significant associations

between higher summary scores and higher population density, higher GDP per capita, a non-

sovereign political status, higher control of corruption and lower religious fractionalization. Of the

included country characteristics, sovereignty status was most strongly associated to the summary

score: it explained 47% of the variation, with higher scores for Caribbean states that have remained

politically affiliated to their former colonizer.

The strength of these associations increased when we omitted the best-performing sovereign state

Cuba, and we then also found additional significant associations between higher summary scores

and a shorter post-independence period, higher government effectiveness, better regulatory quality

Figure 1 Health Policy Performance Score, by country Note: For each Caribbean state separately, we calculated the health policy performance summary score by evaluating whether a state fell in the lower, intermediate or upper tertile for a certain indicator, followed by extracting the percentage of scores in the upper tertile from the percentage of scores in the lower tertile. For example, of the 48 indicators that were included for Cuba, 33 indicators (68.8%) fell in the upper tertile and 8 indicators (16.7%) fell in the lower tertile of the Caribbean distribution. Cuba’s summary score was therefore 68.8% - 16.7% = 52.1%.

Chapter 6

118

Table 1 Association between health policy performance summary score and country characteristics Indicator N r r2 P Demographic characteristics Population (in numbers) 16 0.35 0.12 0.18 Population density (in km2) 16 0.58 0.33 0.02 Economical characteristics GDP per capita (current US$) 15 0.59 0.34 0.02 Health system characteristics Total health expenditure as % of the GDP 11 0.21 0.45 0.53 Total expenditure on health per capita at Purchasing Power Parity (NCU per US$) 11 0.02 <0.01 0.95

General government expenditure on health per capita Purchasing Power Parity (NCU per US$) 11 0.17 0.03 0.63

General government expenditure on health as % of GDP

11 0.60 0.35 0.05

Governance characteristics Current sovereignty status 16 0.66 0.47 <0.01 Years since independence 16 -0.19 0.04 0.48 Government Effectiveness 15 0.44 0.19 0.10 Voice and Accountability 15 -0.25 0.06 0.37 Political Stability and Absence of Violence/Terrorism 15 0.23 0.05 0.42

Regulatory Quality 15 0.28 0.08 0.30 Rule of Law 15 0.30 0.09 0.28 Control of Corruption 15 0.54 0.29 0.04 Cultural characteristics Ethnic fractionalization index 11 -0.44 0.20 0.17 Language fractionalization index 16 -0.42 0.17 0.11 Religion fractionalization index 16 -0.64 0.42 <0.01 Data sources: -Population (in numbers), population density (in km2), GDP per capita (current US$), 2000, were extracted from the United Nations database (www.data.un.org) -Total health expenditure as % of the GDP, total expenditure on health per capita at Purchasing Power Parity (NCU per US$), general government expenditure on health per capita Purchasing Power Parity (NCU per US$), general government expenditure on health as % of GDP, 2004, were extracted from the WHO Global Health Expenditures database (http://apps.who.int/nha/database) -Current political sovereignty (1=sovereign, 2= affiliated) and years since independence, 2010, were derived from (15). -Government Effectiveness, Voice and Accountability, Political Stability and Absence of Violence/Terrorism, Regulatory Quality, Rule of Law, Control of Corruption (ranges from -2.5 (weak) to 2.5 (strong) governance performance), 2000, were extracted from the Worldwide Governance Indicators database (http://info.worldbank.org/governance/wgi/#home) -Ethnic fractionalization index, language fractionalization index, religion fractionalization index (ranges from 0 (low) to 1 (high) diversity), 2001, were extracted from (16).


119

and rule of law and lower ethnic fractionalization (Appendix table A5). We performed a similar

analysis on three partial summary scores for policy implementation, intermediate and final

outcomes separately, but did not find consistent differences between partial summary scores in

their associations with country characteristics, either with (Appendix table A6), or without (results

not shown) the inclusion of Cuba.

Calculation potential health gains

Table 2 depicts, per policy area, the number of observed deaths, the PYLL and the number of

potentially avoidable deaths if all Caribbean states had the age-standardized mortality rates of the

best-performer Martinique. Overall, an annual average of 97,480 men and 82,166 women died in

the 16 included Caribbean states. For the selected causes, the largest number of deaths were found

for cerebrovascular disease for both genders. The principal cause of lost years of Caribbean men,

by far, is homicide/assault. For women, the main cause of lost years is diabetes, closely followed

by breast cancer.

If the mortality rates of Martinique applied in the 15 other Caribbean states, 11,492 male and 9,758

female deaths for the included causes of death could have been avoided annually. For men, the

largest gains would be attained for lung cancer and cerebrovascular disease. For women, it would

be cerebrovascular disease and diabetes. The potential for mortality reduction varies greatly

between Caribbean states (figure 2). The largest relative reductions in total mortality are found in

Belize, where one-third (32%) of the total number of deaths could potentially be avoided. The

largest absolute reduction is observed in the country with the largest population, Cuba. Overall

and in the Caribbean as a whole, the total number of deaths could have been reduced by 12% for

both men and women.

For alcohol-related diseases among men and accidental injury among women the number of

potentially avoidable deaths is negative. This means that for these specific causes the age-specific

mortality rates of Martinique are higher than in several other Caribbean states. It shows that while

Martinique may have done well in terms of low mortality rates for most causes of death,

improvement for these causes is still attainable.

Cha

pter

6

120

Tab

le 2

Ave

rage

num

ber o

f dea

ths,

PYLL

and

pot

entia

lly a

void

able

deat

hs in

Car

ibbe

an st

ates

(n=

16),

by g

ende

r and

pol

icy a

rea,

ca. 2

013

Hea

lth

pol

icy

area

R

elat

ed c

ause

(s)

of d

eath

M

en

Wom

en

Dea

ths

PY

LL

P

oten

tial

ly

avoi

dab

le

dea

ths1

%

pot

enti

ally

av

oid

able

d

eath

s

Dea

ths

PY

LL

P

oten

tial

ly

avoi

dab

le

dea

ths1

%

pot

enti

ally

av

oid

able

d

eath

s

HIV

/A

IDS

HIV

/AID

S 1,

392

43,9

40

1,11

6 80

75

8 26

,358

59

3 78

C

omm

un

icab

le

dis

ease

Co

mm

unica

ble

dise

ase,

exclu

ding

H

IV/A

IDS

1,31

5 33

,407

52

5 40

79

6 20

,293

14

3 18

Can

cer

scre

enin

g M

align

ant n

eopl

asm

s of t

he b

reas

t -

- -

- 2,

970

37,3

74

80

-

Mali

gnan

t neo

plas

m o

f cer

vix

uter

i -

- -

- 1,

009

17,8

26

644

64

Tob

acco

M

align

ant n

eopl

asm

s tra

chea

, br

onch

us a

nd lu

ng

4,50

1 35

,402

2,

582

57

2,44

0 19

,820

1,

277

52

Fer

tilit

y, p

regn

ancy

an

d c

hild

bir

th

Mat

erna

l mor

talit

y -

- -

- 13

9 6,

493

34

24

Ch

ild H

ealt

h

Acc

iden

tal i

njur

y am

ong

child

ren

aged

0-1

4 14

3 9,

896

21

14

67

4,72

3 -5

2 -

Dia

bet

es

Diab

etes

mell

itus

5,35

0 41

,004

2,

374

44

6,57

1 38

,877

2,

873

44

Hyp

erte

nsi

on

Cere

brov

ascu

lar d

iseas

es (s

troke

) 8,

126

47,9

39

2,54

8 31

8,

481

33,7

05

3,20

7 38

Alc

ohol

A

lcoho

l-rela

ted

dise

ases

, exc

ludi

ng

exte

rnal

caus

es

2,34

6 39

,003

-9

7 -

567

7,27

2 19

1 34

Roa

d S

afet

y Ro

ad tr

affic

acc

iden

ts

1,89

3 62

,277

32

5 17

47

1 15

,972

38

4 82

V

iole

nce

H

omici

de/a

ssau

lt 2,

718

114,

243

2,19

8 81

40

1 15

,693

38

5 96

Tot

al

27,7

83

427,

111

11,4

92

41

24,6

68

244,

406

9,75

8 40

Abb

revi

atio

ns: A

IDS,

Acq

uire

d Im

mun

odef

icie

ncy

Synd

rom

e; H

IV, H

uman

Imm

unod

efic

ienc

y V

irus;

PYLL

, Pot

entia

l yea

rs o

f life

lost

1 The

ann

ual a

vera

ge o

f pot

entia

lly a

void

able

dea

ths w

hen

age-

spec

ific

mor

talit

y ra

tes o

f bes

t-per

form

er M

artin

ique

wou

ld h

ave

appl

ied.


121

Discussion

Limitations

The main strength of our study is that it uses the available data from harmonized databases to

generate comprehensive evidence on health policy performance in a large number of Caribbean

territories. Moreover, we estimate the potential impact of implementation of best-practice policies,

which informs national governments about the consequences of continued inaction and should

stimulate them to do better. Our study also has several limitations. First, the incompleteness of

cause-of-death data in several states, as well as the limited data coverage in terms of number of

states and number of health indicators per state may challenge the representativeness and validity

of our results. For example, we noticed that fewer data on policy implementation and intermediate

outcomes were available for non-sovereign Caribbean states (Appendix table A3). Given these

Western countries’ long-standing tradition in evidence-based policy making in their continental

territories, this is rather surprising. The representativeness of our results for the Caribbean as a

whole, and the validity of our health policy performance estimates for the 16 countries included

in the analysis, would decidedly be better if data coverage were more complete. Second, our study

design does not allow a more rigorous impact evaluation of health policies, for example with

Figure 2 Potential annual reduction in mortality if Martinique’s deaths rates applied for the causes of death related to the 11 policy areas.

Chapter 6

122

methods proposed by the World Bank (17). The effectiveness of these policies, however, has

already been demonstrated (12). Moreover, given the limited research capacity in the Caribbean

region thorough impact evaluation studies may not be feasible (18,19), while the current state of

population health in many Caribbean states urges immediate action (20). A third limitation is that

the small number of states for which data were available needs to be considered in the

interpretation of our analyses of significance and other statistical estimates, and challenges

unraveling the determinants of variations in health policy performance in the Caribbean. A

stepwise multivariate regression analysis with forward selection indicated that population density,

ethnic fractionalization, number of years since independence and rule of law explained 99% of the

variation in the summary score (Appendix table A7). The small number of states did not allow us

to tease out the interrelationships between various country characteristics and a state’s summary

score.

Interpretation

Mackenbach & McKee’s assessment found that variations in population health between European

countries are related to differences in health policy implementation (12,13). Our results suggest

that this applies to Caribbean states as well. We do recognize, however, that implementation of

best-practice health policies only partly explains the variation in population health between states.

Better health outcomes can also reflect differences in medical care or more upstream factors, such

as income inequalities and levels of education.

So why have some Caribbean states been more successful in their pursuit of effective health

policies than others? There were several country characteristics we found to be associated to the

health policy performance summary score: sovereignty status, population density, GDP per capita,

control of corruption and the religious fractionalization index, with the strongest association for

sovereignty status.

The finding that population health is generally poorer in sovereign states than in politically

affiliated Caribbean states has been described previously (15,21,22). One of the factors

contributing to these differences is that sovereign states underwent decolonization, i.e., the

transition from a colony to political independence. Decolonization weakens (colonial)

bureaucracies (23), and consequently challenges a state’s capacity to obtain the necessary resources

to implement policy initiatives (15). In agreement with this, more than a decade after sovereign

Caribbean states committed to address their populations’ non-communicable disease crisis (24),

health policy initiatives were still not implemented (5). Another factor that possibly contributes to


123

differences in health policy performance between sovereign and affiliated states is that the strong

political, legal and (socio-)economic ties between affiliated states and their -richer- administrative

countries may make health policy implementation more feasible, for example through the rapid

diffusion of tools and knowledge (25). Best-performer Martinique and third-best performer

Guadeloupe are both French departments and are considered an integral part of the French

Republic since 1946. Their relatively stable political situation, the full French citizenship rights of

their people, and a strong level of collaboration with continental France may explain their superior

health outcomes. In contrast, Aruba, an autonomous country within the Kingdom of the

Netherlands, whose inhabitants do not benefit from the health, educational and social policies that

apply to Dutch citizens in the Netherlands, has the lowest summary score of the included politically

affiliated states; it performs considerably worse on breast cancer, lung cancer and road traffic

mortality, and on teenage pregnancies. Aruba’s local government is considered responsible for the

initiation, implementation and evaluation of health policies, which may be too ambitious given its

small population of approximately 100,000 inhabitants.

Differences in health policy performance between European countries were attributed to

differences in available resources (‘means’) and differences in the willingness to take action (‘will’).

In other words, while adequate financial resources and functioning institutions must be present to

introduce policies, another requirement is that politicians, policy makers and health professionals

acknowledge that a problem exists and are willing to take action accordingly (12,13). Besides

sovereignty status, a Caribbean state’s pursuit of effective health policies was also related to the

financial means they have available: 41% of the variation in the summary score was explained by

GDP per capita. Nonetheless, some governments are doing more or less than expected based on

their economic development. The Bahamas stands out as an underperformer: it has one of the

highest GDPs per capita in the Caribbean, yet scores in the upper tertile for only 9 out of 50

indicators. The Bahamas covers 30 inhabited islands and is the least densely populated territories.

Low population density was related to poorer health policy performance in our analyses, which

may reflect an impaired coverage of (public) health services (26). In comparison, Cuba stands out

as an overachiever in comparison with Caribbean states with a simil ar level of income. Cuba’s low

economic development and relatively exceptional health outcomes challenge the idea that high

economic development is a prerequisite for good population health (27). Moreover, Cuba’s

counterfactual example also shows that relatively weak scores on governance performance

measures such as government effectiveness, voice and accountability, regulatory quality and rule

of law do not exclude successful health policy implementation. Arguably, these findings may

indicate that the “will” to implement policies is fundamental, and that the “means” required will

Chapter 6

124

then follow. Thus, Cuba’s example may illustrate that the link between population health and

economic development can to some extent be broken when engaged technocrats, politicians and

civic leaders are committed to identifying health problems and thinking through solutions.

In the European context, the “will” to implement effective health policies increases when a

population’s cultural attitude moves from traditional survival values, emphasizing economic and

physical security, towards modern self-expression values, emphasizing rising demands of

participation in decision-making in economic and political life (28). For Caribbean states, data on

social values is limited. Nevertheless, our results show that where control of corruption was low,

health policy performance was low as well (29). Corrupted practices are encouraged in collectivist

societies where the focus is on relationships instead of tasks, and where the population accepts

that power is distributed unequally (30,31). We also found that health policy performance is lower

in states that are more religiously fragmented. Just as ethnic heterogeneity in European countries

(13), greater religious pluralism may affect health policy performance by lowering social cohesion

between people of different groups. Dominant groups in religiously pluralistic states may be less

willing to invest in public goods that benefit the whole population, instead of their group in

particular. Another possible explanation is that more religious societies are less inclined to invest

in policies that are incompatible with their religious norms, such as policies for HIV/AIDS and

reproductive health.

Public Health Implications

Our results suggest that differences in health outcomes between Caribbean states are partly due to

variations in the implementation of effective health policies. The calculation of excess deaths

compared to best-performer Martinique shows the potential for considerable health gains in the

Caribbean: an overall mortality reduction in the Caribbean as a whole of 12%. With these results,

we hope to stimulate politicians, policy makers and health professionals around the Caribbean to

come into action and to combine their efforts to improve the health and well-being of their

populations by implementing best-practice health policies. It is, however, likely that these efforts

will remain challenged if the underlying political, economic, and cultural factors are not addressed.

Appendix



125

References

1. Anderson P, Møller L, Galea G. Alcohol in the European Union. Consumption, harm and policy approaches. WHO Regional Office for Europe. Copenhagen: WHO Europe; 2012.

2. Bhutta ZA, Cabral S, Chan CW, Keenan WJ. Reducing maternal, newborn, and infant mortality globally: An integrated action agenda. Int J Gynecol Obstet. 2012;119(Suppl 1):S13–7.

3. International Agency for Research on Cancer. Cervix cancer screening. IARC Handbook of Cancer Prevention volume 10. Lyon; 2005.

4. Lauby-Secretan B, Scoccianti C, Loomis D, Benbrahim-Tallaa L, Bouvard V, Bianchini F, et al. Breast-cancer screening: viewpoint of the IARC working group. N Engl J Med. 2015;372(24):2353–8.


6. Saad PM. Demographic Trends in Latin America and the Caribbean. In: Workshop on demographic change and social policy. Washington D.C.: World Bank; 2009.

7. Luciani S, Cabanes A, Prieto-Lara E, Gawryszewski V. Cervical and female breast cancers in the Americas: current situation and opportunities for action. Bull World Health Organ. 2013;91(9):640–9.

8. Blum RW, Halcón L, Beuhring T, Pate E, Campell-Forrester S, Venema A. Adolescent health in the Caribbean: Risk and protective factors. Am J Public Health. 2003;93(3):456–60.

9. Agozino B, Bowling B, Ward E, St Bernard G. Guns, crime and social order in the West Indies. Criminol Crim Justice. 2009;9(3):287–305.


11. World Health Organization. Analysing mortality levels and causes-of-death (ANACoD) Electronic Tool, Version 2.0. Geneva, Switzerland: Department of Health Statistics and Information Systems, WHO; 2014.

12. Mackenbach JP, Mckee M, editors. Successes and Failures of Health Policy in Europe. Four decades of divergent trends and converging challenges. Buckingham: Open University Press; 2013.

13. Mackenbach JP, McKee M. A comparative analysis of health policy performance in 43 European countries. Eur J Public Health. 2013;23(2):195–201.

14. Hepburn LM, Hemenway D. Firearm availability and homicide: A review of the literature. Aggress Violent Behav. 2004;9(4):417–40.

15. Verstraeten SPA, van Oers HAM, Mackenbach JP. Decolonization and life expectancy in the Caribbean. Soc Sci Med. 2016;170:87–96.

16. Alesina A, Devleeschauwer A, Easterly W, Kurlat S, Wacziarg R. Fractionalization. J Econ Growth. 2003;8(2):155–94.

17. Gertler PJ, Martinez S, Premand P, Rawlings LB, Vermeersch CMJ. Impact Evaluation in Practice. The World Bank/ International Bank for Reconstruction and Development. Washington D.C.: The World Bank; 2011.

Chapter 6

126

18. Cunningham-Myrie C, Reid M, Forrester T. A comparative study of the quality and availability of health information used to facilitate cost burden analysis of diabetes and hypertension in the Caribbean. West Indian Med J. 2008;57(4):383–92.

19. Eldemire-Shearer D. Health information systems in the Caribbean: time for attention. West Indian Med J. 2008;57(4):321–2.

20. Pan American Health Organization (PAHO/WHO). Health in the Americas: Regional Overview and Country Profiles. Washington D.C.: PAHO/WHO; 2017.




24. Pan American Health Organization (PAHO). Public Health Response to Chronic Diseases. In: PAHO/WHO, editor. 26th Pan American Sanitary Conference, 54th session of the Regional Committee. Washington D.C., USA: PAHO/WHO; 2002.

25. Vallin J, Meslé F. Convergences and divergences in mortality. A new approach to health transition. Demogr Res. 2004;10(Suppl.2):11–44.

26. Hanlon M, Burstein R, Masters SH, Zhang R. Exploring the relationship between population density and maternal health coverage. BMC Health Serv Res. 2012;12(416).


28. Mackenbach JP. Cultural values and population health: a quantitative analysis of variations in cultural values, health behaviours and health outcomes among 42 European countries. Heal Place. 2014;28:116–32.

29. Factor R, Kang M. Corruption and population health outcomes: an analysis of data from 133 countries using structural equation modeling. Int J Public Health. 2015;60(6):633–41.

30. Seleim A, Bontis N. The relationship between culture and corruption: a cross-national study. J Intellect Cap. 2009;10(1):165–84.

31. Husted BW. Wealth, Culture, and Corruption. J Int Bus Stud. 1999;30(2):339–59.

Chapter 7

General discussion

Chapter 7

130

This thesis aims to provide a better insight into the health situation of the Dutch Caribbean, and

the factors related to this health situation, in particular the role of the political context and health

policy performance. We have addressed this aim in two research questions. The first research

question seeks to describe the health situation in the Dutch Caribbean through a comparison of

health status in the Dutch Caribbean, the Netherlands, and other Caribbean states. The second

research question investigates the influence of the political context and health policy performance

on population health in Caribbean states. Considering that the implementation of health policies

is an important outcome of the political decision-making process, we also examine how possible

interactions between these two factors influence population health in Caribbean states.

This final chapter summarizes our main findings, and discusses our results in light of their

methodological limitations. Moreover, we interpret our main findings in reference to earlier

studies, and propose ways for improving health research and reducing between-country disparities

within the Kingdom of the Netherlands. We close this chapter with an overall conclusion of this

thesis’ findings.


I. What is the health status of the Dutch Caribbean population?

For the first research question, we sought to explore how the health status of the Dutch Caribbean

population compares to the Dutch population, and to the populations in other Caribbean states.

Comparative health data from the Dutch Caribbean islands of Bonaire, Saba, and St. Eustatius and

St. Maarten were unavailable, making Aruba and Curaçao (the latter being the largest territory of

the former Netherlands Antilles (NA)) the islands of main focus. We considered three different

health measures: self-reported health indicators, life expectancy and cause-specific mortality rates.

Overall, we found that the health status of the Dutch Caribbean population is poorer than that of

the Dutch population (chapter 2 and 3), and that of populations in Caribbean states that are

politically affiliated to France, the United Kingdom (UK), and the United States (US) as well

(chapter 5).

In chapter 2, we made a three-way comparison between the Dutch Caribbean population living

in Curaçao, Dutch Caribbean migrants living in the Netherlands, and non-migrants living in the

Netherlands. The results of chapter 2 indicate that, perhaps surprisingly, the Dutch Caribbean

population reported better physical and mental health than Dutch Caribbean migrants in the

Netherlands during the 2012/2013 period. Health outcomes were poorer for Dutch Caribbean

General discussion

131

migrants compared to non-migrants living in the Netherlands as well. These results persisted even

when controlling for sociodemographics (age, marital status, education and employment), risk

factors (tobacco and alcohol use, overweight and physical activity) and urbanization (in the

Netherlands). The exceptions were overweight/obesity and diabetes; prevalence rates for

overweight/obesity (men only) and diabetes (both genders) were similar for the Dutch Caribbean

populations living in Curaçao and in the Netherlands, and higher than in Dutch non-migrants. For

Aruba, no comparative data were available to include in this study.

The study in chapter 3 demonstrates that levels of amenable mortality, deaths from causes that

are considered unnecessary in the presence of timely and effective healthcare, have been

consistently higher in Aruba and the NA/Curaçao than in the Netherlands during the 1988-2014

period. If mortality from amenable causes were reduced to similar levels as in the Netherlands,

men and women in Aruba would add respectively 1.19 years and 0.72 years to their life

expectancies. In Curaçao, this would be 2.06 years and 2.33 years. If mortality from amenable

circulatory causes (which includes ischemic heart disease, cerebrovascular disease, and

hypertensive disease) were reduced to similar levels as in the Netherlands, men and women in

Aruba would add respectively 1.02 years and 0.56 years to their life expectancies. In the

NA/Curaçao, this would be 0.65 years and 0.55 years. Perinatal deaths are an important additional

contributing cause-of-death in the NA/Curaçao, and are responsible for 0.79 years (men) and 0.90

years (women) of the total mortality gap with the Netherlands. Nephritis/nephrosis is also

important for the NA/Curaçao, where it contributes 0.12 years for men and 0.21 years for women.

In addition, reduction in breast cancer mortality to the level of the Netherlands would add 0.22

years in Aruba and 0.21 years in the NA/Curaçao to female life expectancy.

The results of chapter 5 show that Aruba and NA/ Curaçao experienced declines in male and

stagnation in female life expectancy, and that life expectancy trends have diverged relative to the

Netherlands during the 1980-2014 period. In contrast, life expectancy trends in Caribbean

dependencies that had relatively low levels of autonomy, Martinique, Guadeloupe and the UK

islands, ran in parallel with those in their administrator countries. Decomposition of life expectancy

differences between Western countries and their dependencies point to the important contribution

of mortality from cardiovascular diseases and external causes to the excess mortality in the

Caribbean states. The magnitude of mortality differences with Western countries, however, differs

greatly. Unlike other affiliated states, Dutch dependencies did not experience substantial

improvements in cardiovascular mortality during the 1980-2014 period. Also, the NA/Curaçao

experienced sharp rises in homicide during the study period, alongside the US territories Puerto

Chapter 7

132

Rico and the US Virgin Islands. In addition, transport accidents are an important additional

contributor to higher external cause mortality in NA/Curaçao, and the main contributor to the

high external causes mortality in Aruba.

II. To what extent do differences in population health in the Caribbean reflect differences in political context and health policy performance?

For the second research question, we turn to the role of political context and health policy

performance as potential factors that influence population health variations between Caribbean

states. We considered three determinants for the political context: current political sovereignty

status, decolonization and degree of political affiliation. Overall, we found that decolonization has

increased the life expectancy gap between currently sovereign and affiliated Caribbean states, and

that a strong political affiliation to a Western country is positively related to population health and

health policy performance in Caribbean states.

In chapter 4, we assessed how disparities in life expectancy between currently sovereign and

affiliated Caribbean states have developed over time. We saw that life expectancies in currently

sovereign states were already lower than in currently affiliated states before the decolonization

events in the last half of the 20th century. Decolonization, the process in which a colony becomes

politically independent, coincided with reductions in life expectancy growth in the decade after

independence, which did not correspond with similar changes in economic performance. After the

decolonization events, the gap in life expectancy between currently sovereign and affiliated states

accelerated in the 1990's, and it continues to increase to the present-day.

In chapter 5, we investigated whether differences in life expectancy trends between Caribbean

dependencies and their Western administrators were related to their degree of political

independence. We found that Caribbean states with a strong political affiliation to their Western

administrators, Martinique, Guadeloupe, and the UK islands, experienced more favorable life

expectancy trends during the 1980-2014 period than Caribbean states with higher levels of

autonomy, the US Virgin Islands, Puerto Rico, Aruba and Curaçao. In comparison with the

Western administrator, the underlying trends in causes-of-death are similar among Caribbean

states and possibly reflect region-specific drivers. The politically affiliated Caribbean states,

however, differ greatly in the magnitude of changes in mortality, most notably for cardiovascular

diseases and external causes. The least favorable trends are found for affiliated states with a high

degree of political autonomy.

General discussion

133

In chapter 6 of this thesis, we determined whether and why 16 Caribbean states vary in health

policy performance in 11 different policy areas related to regional health objectives of the Pan

American Health Organization (PAHO) during the 2010-2015 period: 1. HIV/AIDS, 2.

Communicable disease, 3. Cancer screening, 4. Tobacco, 5. Fertility, pregnancy and childbirth, 6.

Child health, 7. Diabetes, 8. Hypertension, 9. Alcohol, 10. Road safety, and 11. Violence. Our

findings suggest that variations in population health outcomes between Caribbean states are related

to differences in the implementation of effective health policies. In other words, Caribbean states

with better health outcomes were more likely to have implemented a comprehensive set of best

practice health policies for the 11 policy areas included in this study. Martinique, Cuba and

Guadeloupe had the highest health policy performance summary scores, and Guyana, Belize and

Suriname had the lowest. Aruba has the lowest health policy performance summary score of five

politically affiliated Caribbean states included in this study, indicating that performance in 11 policy

areas related to important regional health objectives is suboptimal compared to other politically

affiliated Caribbean states. In terms of health outcomes, Aruba performs considerably worse on

breast cancer, lung cancer and road traffic mortality, and on teenage pregnancies. The latest

available mortality data from Curaçao dated from 2007 and we therefore had to refrain from doing

a similar analysis for Curaçao.

Again, we found a strong association with political sovereignty status: Caribbean states that

remained politically affiliated to Western countries had generally higher health policy performance

summary scores. Health policy performance was also better in states where the population density,

GDP per capita, and control of corruption is high, and states that are less religiously fragmented.

The small number of states did not allow us to tease out the interrelationships between the various

country characteristics and a state’s health policy performance summary score. In this analysis,

Cuba stands out as an overachiever considering its political independence, its low economic

development and its exceptional health outcomes. The potential for mortality reduction from

causes that are avoidable through health policy interventions varies greatly between Caribbean

states, and is considerable for the Caribbean as a whole: if the mortality rates of Martinique applied

in other Caribbean states, an overall mortality reduction of 12% could be obtained.

Methodological considerations

The studies in this thesis used data derived from cross-sectional health surveys, mortality

registration systems, harmonized international databases, and country reports to compare

population health in the Dutch Caribbean with the Netherlands and with other Caribbean states.

Chapter 7

134

In this section, we will critically discuss data considerations and research designs, and their

potential impact on the representativeness and validity of our results.

Descriptive comparative research

A thorough description of the health patterns across territories is a necessary first step towards

explaining potential causes for the variations between territories. For the first research question,

we therefore focused on comparing the health status of the Dutch Caribbean population in an

international context. We concluded that the health status of the Dutch Caribbean population is

generally poorer than that of the Dutch population, and that of populations in other politically

affiliated Caribbean states as well. We noticed, however, that few data from mortality and cancer

registrations, health surveys, infectious disease surveillance and administrative systems were

available in all Dutch Caribbean territories. Given the Kingdom’s longstanding tradition of

evidence-based policy making in the Netherlands, this is rather surprising. For some common

health measures, data are not collected or not published on a regular basis, or this has only recently

started as in the case of Bonaire, Saba and St. Eustatius. For other health measures, owners seem

unwilling or unable to share the data from their organization. This has restricted our current

account on the health status in the Dutch Caribbean in two important ways.

First, we were not able to include data from all Dutch Caribbean territories in our studies. As

comparative data were not available, Bonaire, Saba, St. Eustatius and St. Maarten were excluded

from our analyses. Moreover, in some studies we had to exclude either Aruba or Curaçao, or had

to restrict our analysis of trend data. For example, no recent health survey data were available from

Aruba for our study in chapter 2. In Curaçao, the publication of mortality statistics lags behind,

with the latest data from 2007. Our comparison of amenable mortality in chapter 3 was therefore

based on the latest time-period for which data for the Netherlands, Aruba and Curaçao were

available, i.e. 2005-2007, which raises obvious questions for present day relevance as rates may

have changed over time. Moreover, mortality data from Curaçao were considered too old for the

study on health policy performance in chapter 6, as its measurement preceded data on the

implementation of health policies and on intermediate health outcomes. Although we were able

to study the health status of the two largest territories in the Dutch Caribbean, Aruba and Curaçao,

the representativeness of our results for the Dutch Caribbean as a whole would decidedly be better

if health data coverage were more complete. Recently, health data from Bonaire, Saba, and St.

Eustatius have become available. Life expectancy in the three special municipalities of the

Netherlands was a year shorter than in the European part of the Netherlands (1), and self-reported

rates for obesity and diabetes vary considerably between the islands, but are much higher than in

General discussion

135

the Netherlands (2). We therefore have no reason to believe that the health situation on the other

Dutch Caribbean islands is much better than that on Aruba and Curaçao.

Second, the majority of our descriptive analyses were based on life expectancy and mortality data.

To obtain a more complete picture, we compared self-reported health status and risk-factors

between Curaçao and the Netherlands (chapter 2), and have complemented our analyses with data

from country reports and harmonized databases. Nonetheless, if we had been able to assess health

data from sources that shed a broader light on morbidity, such as mental health, infectious disease

and cancer incidence, this would have permitted a more complete view of the health situation in

the Dutch Caribbean. For example, there are indications that for certain aspects of population

health, health outcomes in the Dutch Caribbean are more favorable than in the Netherlands. Self-

reports of feelings of sadness/depression and migraine are less common in Curaçao than in the

Netherlands, and perhaps relatedly, Curaçaoans less frequently report (problematic) alcohol use

(3). Moreover, in line with the lower rates of lung cancer mortality, tobacco use is less prevalent in

Curaçao than in the Netherlands as well (3). Regardless of these counterexamples, we find it

unlikely that this would affect the robustness of our overall conclusion that population health in

the Dutch Caribbean is relatively poor as compared to the Netherlands and other politically

affiliated Caribbean states. Measuring the final outcome of life, our comparative (trend) analyses

on mortality data give insight into many diseases and external factors that influence population

health in Aruba and Curaçao, and a large number of specific cause-of-death rates are higher than

in the Netherlands.

Explanatory comparative research

For the second research question, we attempted to explain the potential factors contributing to

the health variations between Caribbean states, in particular political context and health policy

performance. Our use of observational study designs for explanatory analyses may raise questions

on the validity of our results.

First, the use of observational data to determine relationships between population health and

potential determinants across territories is not without its concerns. In health research, clinical

studies in particular, the gold standard to provide evidence on a causal relationship between

exposure and outcome remains the randomized control trial, although this is now slowly changing

(4,5). In contrast to experimental studies, in observational studies the exposure to a determinant

or intervention is not manipulated by the researcher. In observational studies, it is often impossible

to completely eliminate the bias of selective exposure to a determinant of interest and that of

Chapter 7

136

confounding variables (i.e., factors that independently affect exposure and health outcome). The

main methodological limitation of our studies is therefore that they are not able to provide strong

evidence on cause and effect (6). This is, however, a direct consequence of the nature of the

exposure variables we investigated, i.e., the political context and health policy implementation at

the national level, which cannot be randomized. While this is an important disadvantage, an

important advantage of observational studies is that they give insight into real world conditions,

and do not require a reduction of the multifaceted complexity underlying disease, injury and death

(7).

Second, the studies in chapter 4 and 5 indicate that a strong affiliation to a Western country and

hence, lower levels of political autonomy, is an important predictor for favorable population health

outcomes in the Caribbean. This finding is in accordance with the findings of previous studies we

discussed in the introduction (8–10). The association between political sovereignty and population

health also fulfills several Bradford Hill criteria for potential causality (11). One, the association is

strong: there is almost clear-cut division in political sovereignty status for life expectancy and

economic development (see figure 3, chapter 1). Two, the temporal relationship between political

sovereignty and population health is consistent with what one would expect if changes in political

autonomy influenced health: decolonization in the last half on the 20th century preceded worsening

of life expectancy growth in currently sovereign Caribbean states (chapter 4). Three, we observed

a gradient of political dependency and mortality outcomes, reminiscent of a dose-response

relationship: life expectancy trends in political affiliated states with lower levels of political

autonomy are more favorable than in affiliated states with higher levels of political autonomy

(chapter 5). Four, the findings from the Caribbean are consistent with observations from the

Pacific region: relative to sovereign Pacific states, economic and health outcomes are also more

favorable in politically affiliated island states (8,10). Five, plausible explanations are available for

the link between political sovereignty and population health based on our studies in chapter 4 and

5. On the one hand, a stronger political affiliation can increase a government’s capacity to

influence areas that contribute to population health, for example through its bureaucracy’s ability

to implement and enforce effective policies. On the other hand, a strong political, cultural and

(socio)economic relationship with a prosperous administrator country will more likely facilitate the

diffusion of knowledge, tools and finance necessary to implement effective policies. Population

health in the territories of the French islands, for example, benefit from strong collaborations with

metropolitan institutes and from the European Union’s support for its economic, social and

cultural development through structural funds.

General discussion

137

The assumption of a causal relationship between political sovereignty and health, however, needs

further scrutiny, as there may be other country characteristics that independently affect the

magnitude of health variations across Caribbean territories (confounders). One obvious candidate

is economic performance. In his analysis of 67 island states, Bertram showed that the gap in

economic performance between currently affiliated and sovereign states was well established prior

to the first wave of decolonization in the 1950’s (10). The petition for independence during the

last half of the 20th century was primarily initiated by the local governments of colonized countries

and territories in the Caribbean (12,13), which indicates that poorer colonies were more likely to

opt for independence. Given the strong association of population health with economic

development (14), economic conditions may have plausibly contributed to both current political

status and population health outcomes of Caribbean states.

We consider population health, however, unlikely to be dictated solely by economic performance,

as we previously mentioned in chapter 1. This is perhaps best illustrated by the counterfactual

experience of Cuba. With its low economic performance, but exceptionally good health outcomes

and health policy performance, Cuba stands out as an overachiever in comparison to Caribbean

states with a similar level of income (chapter 6). Also notable is that we did not find an indication

that economic conditions have influenced mortality developments in chapter 4 and 5. In chapter

4, we observed that reductions in life expectancy growth in the decade after independence occurred

irrespective of changes in economic performance. In chapter 5, while life expectancy trends have

been least favorable in the Dutch and US territories, we did not find reduced economic growth in

these territories relative to France and United Kingdom during the 1980-2014 period. This

suggests that population health in the Caribbean is also partly determined by other factors, perhaps

the prioritization of health and education on a national level (15). Importantly, Cuba’s example

illustrates that the dependency of population health on economic performance can to some extent

be broken when politicians, policy makers and health professionals are committed to identifying

health problems and thinking through solutions (16).

Sovereign states may also differ from states that are politically affiliated in other respects than

political sovereignty and economic performance. Current population health reflects the

accumulation of complex historical interactions between economic, social and other health-

influencing factors, but quantitative data for these historical factors are rarely available. We

considered the identity of the former colonizer and the duration of colonization as potential

confounders, as they were previously shown to be important predictors for contemporary

outcomes of infant mortality and economic performance (17). Although the majority of currently

Chapter 7

138

sovereign states in the Caribbean are former British colonies, we find it unlikely that the identity

of the former colonizer and the duration of colonization are responsible for the current health

divide between currently sovereign and affiliated Caribbean states. Life expectancies in currently

affiliated UK islands are more favorable than currently sovereign, former UK colonies, as well.

Moreover, we consider differences in the length of colonization as marginal, as Caribbean states

were at least 250 years under colonial rule.

Third, in chapter 6 we empirically evaluated one of the main outputs of the political decision-

making process on health: the implementation of health policies. Our results show that differences

in health outcomes between Caribbean states are partly attributable to variations in the

implementation of effective health policies. A comparable analysis of 43 European countries

reached a similar conclusion (18,19). A major strength of this study was that it was able to assess

the potential health gains of best practice policies in a region with low research capacity, which

reduces the feasibility for more comprehensive policy impact evaluation studies. The great reliance

of this cross-sectional analysis on data of undetermined validity from harmonized databases,

however, may affect our findings if health policy performance is also related to the quality and

availability of health data. For each Caribbean state, we therefore applied a rigorous procedure to

determine whether quality and completeness of death registration data was suitable for inclusion

in our analysis. For data on policy implementation and intermediate outcomes, however, we

noticed that fewer data were available for Caribbean states that are politically affiliated to Western

countries. The main reason for this is that international organizations such as the World Health

Organization and the United Nations do not include these territories in their databases. Surely, we

would have obtained more reliable estimates of health policy performance if data coverage had

been more complete. Nonetheless, we do not expect that this has strongly influenced our health

policy performance rating: based on data from available countries, we were able to show that the

correlations between indicators for policy implementation, intermediate and final health outcomes

are largely in accordance with what one would expect if policy implementation induces changes in

intermediate outcomes, and if changes in intermediate outcomes induce changes in final outcomes.

Another limitation of this study is the relatively small number of states (16) included in this study,

which reduced the statistical power of our analyses, and restricted further explorations of the

interrelationships between various country characteristics and a state’s health policy performance

summary score.

General discussion

139

Interpretation of findings

Health status of the Dutch Caribbean population in Aruba and Curaçao

A major strength of the mortality data we analyzed in this thesis is that it allows for cross-territory

comparisons to benchmark local health outcomes in an international context. This section

compares the mortality rates from causes-of-death that are potentially avoidable through

healthcare and/or policy interventions. In other words, the mortality rates reflect premature deaths

from selected conditions that are considered either treatable through timely and effective health

care (amenable mortality), or preventable by the implementation of effective public health

interventions (preventable mortality). Because the degree to which death from certain diseases can

be prevented becomes progressively more doubtful at older ages, age limits were set for non-

external mortality causes in line with other research (20).

Figure 1 depicts an overview of the avoidable mortality rates in the Dutch Caribbean islands of

Aruba and Curaçao, the Netherlands, and the French islands of Guadeloupe and Martinique. We

included up to five years of the most recent data during the 2010–2014 period for each Caribbean

state, with the exception of Curaçao (2005-2007). For each territory, age-standardized mortality

rates per cause-of-death are given. The results indicate that people in Aruba and Curaçao were

more likely to die from breast cancer, cervix cancer, ischemic heart disease, cerebrovascular

disease, pneumonia, nephrosis/nephritis and transport accidents than in the Netherlands,

Guadeloupe and Martinique. In Curaçao, important additional contributors of excess mortality are

HIV/AIDS, cancer of the colon and rectum, diabetes, hypertensive heart disease, perinatal deaths

and violence. Mortality rates from lung cancer are lower in the Dutch Caribbean than in the

Netherlands, possibly because tobacco use is more prevalent in the Netherlands (3).

Chapter 7

140

Figure 1 Age-standardardized mortality rates for selected avoidable causes of death, latest data available

General discussion

141

Avoidable mortality is a function of the underlying disease incidence, access to high quality

healthcare, and implementation of effective health policies. Disease incidence is also influenced by

factors outside of the health system, such as socioeconomic conditions, health behaviors and

environmental factors. Many of these factors are less favorable in Aruba and Curaçao than in the

Netherlands, probably because policy responses are still lacking. For example, income distributions

are far more unequal, the proportion of people living in poverty is higher, and educational

attainment is lower (21,22). So far, there is no firm commitment from the local governments to

carry through interventions that would reduce socioeconomic inequalities within their societies,

such as policies that encourage asset-building for working families, improve educational outcomes

and prevent housing segregation (23). Obesity is twice as prevalent in Curaçao than in the

Netherlands, and sport-participation almost twice as low (3). Relatedly, there is a high density of

fast food outlets, a scarcity of fresh produce in neighborhood stores and an unsafe infrastructure

for walking and cycling in the islands, factors that have been previously shown to be associated to

higher risks of obesity (24–26). Also, SO2-emissions recorded in Curaçao are among the highest in

the world (27). In the poor neighborhoods that are affected by the air pollution this has regularly

led to serious acute health effects such as coughing, vomiting and headaches (28), and the self-

reported prevalence of asthma in these neighborhoods is higher than in other poor neighborhoods

as well (29). The efforts and activities taken by the local government to prevent these health effects

in the future are, however, not yet sufficiently visible to the public. Although it is difficult to

ascertain how much these social determinants of health contribute to the high rates of avoidable

mortality in the Dutch Caribbean, observed differences in avoidable mortality in cross-country

comparisons, however, cannot be solely attributed to differences in disease incidence (30,31).

Therefore, while policies are needed to address the social determinants of disease, an important

aspect contributing to the higher rates in avoidable mortality is that governments in the Dutch

Caribbean have, so far, not optimally addressed their population’s health needs.

There are indications that the higher rates of avoidable mortality reflect that the provision of

curative and preventive health services, and health policy performance in general, falls short in

Aruba and Curaçao. In the study in chapter 3, we already mentioned that the access to healthcare

services is inadequate and that the (implementation of) legislation that regulates the quality of

healthcare services lags behind. The -often underreported- mortality cause “misadventures during

medical care” reflects the number of deaths from injuries caused by an adverse event during

medical treatment. Under the assumption that this cause-of-death is not more frequently

underreported in the Netherlands than in the Dutch Caribbean, the mortality rate for this cause is

still 15 (Curaçao) to 21 (Aruba) times higher in the Dutch Caribbean (own calculations). The access

Chapter 7

142

and quality of healthcare services is likely to vary greatly among health professionals, types of care

and healthcare organizations. Nonetheless, the negative experiences of encounters in healthcare

that were published in the (social) media reveal that in the eyes of some patients, professional

standards are regularly not met by health professionals on the islands. In Curaçao, for example,

where over the years several specialists were considered unfit to provide medical care by their

colleagues in the St. Elisabeth Hospital, yet could continue their medical practice elsewhere on the

island. Evidently, much of the concern on the quality of healthcare services, and the self-regulating

capacity of the healthcare system, has developed through these publically discussed events.

Beyond these general features of the health system, the study in chapter 3 also revealed that there

is much room for improvement in the healthcare processes that underlie deaths that are amenable

to healthcare, circulatory diseases, breast cancer, perinatal deaths and nephritis/nephrosis (the last

two causes solely in NA/Curaçao) specifically. We also discussed that Curaçao struggles with

providing sufficient coverage in areas of preventive care, such as cervix cancer, colon cancer and

(vaccine-preventable) infectious diseases. Another policy area that stands out is that of sexual

health. Teenage pregnancies are relatively high in both Aruba and Curaçao (chapter 3), and

HIV/AIDS mortality in the latter territory (figure 1), which suggests that increased attention to

planning and implementation of effective strategies are required. So far, the causes that contribute

to the high mortality for external causes have also received unsatisfactory attention in the health

policy priorities set in Aruba and Curaçao. While the economic cost of the many motor vehicle

crashes is regularly discussed in the media, traffic safety is generally not considered a public health

concern. One viewpoint contributing to this is that traffic accidents are often considered avoidable

by becoming a “better” driver, rather than the anticipatable result of a lack of regulation and

enforcement in road safety policy. Moreover, the high rates of homicide mortality in Curaçao are

among the highest in the world, but have not been declared a public health crisis. Violent deaths

mainly affect young men from the lower social classes, a group already characterized by high drop-

out rates and youth unemployment (32). Regardless of the victims and perpetrators frequently

alleged links to organized crime, however, the toll of trauma affects the well-being and safety of

those witnessing the violent events as well.

In the absence of more detailed information on morbidity, the causes-of-death we mentioned

above are therefore likely candidates for targeted interventions. The studies in chapter 3 and 6 of

this thesis suggest that considerable health gains could be achieved in Aruba and Curaçao if these

causes-of-death would be targeted with best practice policies. To illustrate, the calculation of excess

deaths compared with best-performer Martinique estimates the potential for mortality reduction

General discussion

143

from the implementation of best-practice policies in 11 policy areas (chapter 6). If the mortality

rates of Martinique applied in Aruba, an average of 121 deaths could have been avoided annually

during the 2011-2014 period, an overall mortality reduction of 14%. The potential for mortality

reduction in Curaçao would be 11% of total mortality, equating an average of 122 deaths annually

during the 2005-2007 period.

It should be noted that the retrospective character of our studies largely exclude the effects of the

current political crisis in Venezuela on population health. The resurgence of vector-borne and

vaccine-preventable disease in Venezuela (33,34), the impact of this crisis on the islands’

economies, the influx of relatively large numbers of refugees, in addition to the austerity measures

implemented to achieve sustainable public financial management, places additional stress on the

islands’ health systems. Possibly, population health outcomes in Aruba and Curaçao have further

deteriorated analogous to the effects of financial crisis in other countries (35,36). In light of the

slow pace of improvements that the healthcare systems in Aruba and Curaçao have experienced

in the past decade, and the undesirability of allowing the health situation to further deteriorate

relative to other Caribbean states, more efforts are needed. The need to do so is best captured in

the words of Richard Horton, who states that the “health system is the most visible and tangible

expression of a society’s concern for the welfare and wellbeing of its citizens”.

Health status of the Dutch Caribbean population in the Netherlands

Higher rates of obesity, diabetes and hypertension among the Dutch Caribbean population in the

Netherlands relative to the Dutch population have previously been documented (37–40). Where

these health outcomes corresponded with the analysis in chapter 2, we came to similar

conclusions. The study in chapter 2 goes beyond earlier studies by adding the ‘home-country

perspective’ and by considering a comprehensive set of health outcomes and health behaviors. We

expected that the health outcomes and behaviors of Dutch Caribbean population in Curaçao

would be poorer than that of the Dutch Caribbean population in the Netherlands. The reason for

this is twofold. First, the Netherlands is a country known for its quality healthcare services and its

comprehensive implementations of public health interventions. Second, many young adults leave

the islands to continue their education and further their careers in the Netherlands, indicated by

the popularly used term “brain drain”. In agreement with this, we observed that, compared to the

Dutch Caribbean population in Curaçao, the Dutch Caribbean population in the Netherlands is

younger and more highly educated. Surprisingly, however, we found that Dutch Caribbean

migrants living in the Netherlands generally reported poorer physical and mental health than both

Chapter 7

144

Dutch Caribbean non-migrants in Curaçao and Dutch non-migrants in the Netherlands, even

when controlling for sociodemographic variables such as age and educational level.

Although poorer health outcomes among the Dutch Caribbean population in the Netherlands

have been commonly reported, the explanations for these findings have remained rather

speculative. The prevailing explanations typically emphasize the role of individual-level factors and

their influence on unhealthier health behaviors. For example, the higher incidence of sexual

transmitted infections (STI’s) among the Dutch Caribbean population in the Netherlands were

believed to reflect an inclination for high-risk sexual behaviors (41), but the evidence base for this

relationship was later contested (42). Moreover, excess diabetes mortality among Dutch Caribbean

migrants in the Netherlands was previously related to their relocation to a more prosperous

country, which was suggested to lead to drastic lifestyle changes, such as declines in physical

activity and high-fat, energy-dense diets (43). Such individual-level explanations for poorer health

outcomes are not limited to the Netherlands. In Curaçao for example, where traits that reflect a

lack of character have at times been proposed by health professionals as an explanation for the

relatively high obesity prevalence (44). The logic underlying such individual-level explanations can

be problematic: it may essentially lead to the perception of a stereotypical, homogenized

migrant/ethnic population (45–48), which is a gross oversimplification of the diversity in ethnicity,

religion, language, and socio-economic class in the Dutch Caribbean. In addition, the migration

turnover is relatively high in the Dutch Caribbean, illustrated by the fact that 33% of the population

in Aruba and 24% of the population in Curaçao were born abroad, mainly in the Netherlands,

Colombia, the Dominican Republic and Venezuela (49,50). Although certain cultural

characteristics are commonly equated with a population living in a certain territory, the evidence

base for this is weak (51).

The results in this thesis provide an alternative explanation for the higher rates of obesity and

related conditions, and the higher incidence of STIs, relative to the Dutch population. Prevalence

rates for diabetes and overweight/obesity were generally similar for the Dutch Caribbean

populations in Curaçao and in the Netherlands (chapter 2), and rates of HIV/AIDS mortality

were also relatively high in the Dutch Caribbean islands (chapter 5). The higher prevalence of

obesity and relatedly diabetes, and STIs, among the Dutch Caribbean population in the

Netherlands therefore reflect their migration from an area in which the underlying risk factors are

more prevalent, partly because the governments have not properly invested in interventions to

prevent, detect and treat these conditions (chapter 3 and 6). We previously mentioned that in

Curaçao, there is a high density of fast food outlets, a scarcity of fresh produce in neighborhood

General discussion

145

stores and an unsafe infrastructure for walking and cycling, environmental factors that likely

contribute to higher risks of obesity (24–26). In turn, the heightened risk for STIs may reflect that

the procedures necessary to quickly detect and respond to infectious disease outbreaks are

suboptimal (52), as well as the investments in sexual health prevention.

While the health situation in the home-country of migrants may therefore play a role in shaping

certain health outcomes, our study in chapter 2 concluded that the overall poorer physical and

mental health status of Dutch Caribbean migrants in the Netherlands is suggestive of host-country

specific stressors. Based on the discussion of recent studies on migrant health in the Netherlands

and previous health literature, we proposed three host-country-speci c stressors that affect the

Dutch Caribbean population more than the Dutch population in the Netherlands: perceived

discrimination, spatial concentration in multi-ethnic neighborhoods and reduced social mobility.

Experiences of stigmatization are, however, not limited to the Netherlands: racial patterns of

deprivation and exclusion stubbornly persists in the Dutch Caribbean as well, where the majority

Afro-Curaçaoan population continues to be associated with the lower classes (53).

In contrast to our findings on health outcomes, we found that risk factor prevalence among Dutch

Caribbean migrants in the Netherlands converges to the Dutch population norm; rates for tobacco

and alcohol use were in-between the lower rates of Dutch Caribbean non-migrants in Curaçao and

the higher rates of Dutch non-migrants in the Netherlands. This indicates that many Dutch

Caribbean migrants adopt the health behaviors and attitudes of Dutch society, which is line with

key indicators of sociocultural integration: Dutch Caribbean migrants frequently spend their free

time with Dutch non-migrants and 45% of the marriages among Dutch Caribbean migrants is with

a native Dutch partner (54). As an emphasis on individual-level explanations for ethnic health

disparities may obscure the impact of host-country-specific factors, we suggest that the ‘home

country’ perspective of migrant health may provide a valuable addition to inform policy makers.

The political context of population health in the Dutch Caribbean

This thesis demonstrates a positive effect of a strong political affiliation to a Western country on

population health in Caribbean states. The idea that a strong affiliation to a Western country largely

outweighs its disadvantages in the modern age should not come as a surprise, as the broad literature

of studies from the fields of political sciences, sociology and economics on this topic suggest (8–

10,55–59). In the words of McElroy and Mahoney, affiliated states “remain unwilling to trade the

visible security, affluence and standard of living of affiliation for the less tangible but more costly

Chapter 7

146

rewards of autonomy”(60). Notwithstanding, the studies in this thesis expand the existing

knowledge in two important ways.

First, we observed that life expectancies in currently sovereign states were already lower than in

currently affiliated states during the colonial period, that decolonization coincided with a reduction

in life expectancy growth, and that the gap in life expectancy between currently sovereign and

affiliated states continues to increase to the present day (chapter 4). Colonial institutes varied

greatly in their promotion of prosperity for the (nonindigenous) public good (inclusive institutes)

or for a small elite (extractive institutes), across colonies and over time (57). Kenny describes that

political decisions during decolonization can weaken the (colonial) bureaucracies that served their

public’s interests through the institutionalization of corruption (61), which consequently may have

disrupted a states’ capacity to solve problems, to provide necessary resources and to implement

effective policy initiatives in areas that contribute to population health. The continuous

involvement of Western countries in currently affiliates states, in contrast, may have plausibly

resulted in the co-evolvement of good governance and state capacity alongside their continental

territories. “Western” influences have been consequential in state building and state capacity in a

multitude of countries, and hence, population health and economic development, probably by

increasing “the scope of their policies as the resources at their disposal increased” (55). Over time,

these institutional effects have cumulated into divergent development trajectories (62), which

underlie the gap in economic performance and population health between politically sovereign and

affiliated Caribbean states to the present day.

Second, the comparison of life expectancies and cause-specific mortality rates across politically

affiliated states suggests that it is not the division in political sovereignty in itself that dictates the

gap between affiliated and sovereign Caribbean states, but rather its reflection of a “gradient of

political autonomy” (chapter 5). This “gradient of autonomy” also translates into a variety of

governance structures on the area of public health: governmental (health) organizations in

Caribbean states with a lower degree of autonomy have more and stronger collaborations with

metropolitan (health) institutes. In other words, the strong political, cultural, and (socio)economic

relationship with a Western country may (eventually) result in the availability of resources (e.g.

human capital, financial resources and investments, and technology) that contribute to better

population health. For example, health system governance on the French islands, which are

considered an integral part of the French Republic, is implemented with oversight from the

Ministry of Health in metropolitan France, since 2010 by the “Agence Régionale de Santé (63).

Likewise, Caribbean health organizations in the dependencies of the United Kingdom actively

General discussion

147

work together with metropolitan organisations to protect and improve the health of their

populations. While the island organizations are responsible for the implementation of policies, the

European organizations pro-actively support the islands’ governments with the provision of

healthcare and the implementation of public health interventions (64).

To illustrate the situation in the Dutch Caribbean further, we will now view health system

governance in the Dutch Caribbean context. In contrast to other Western administrators, the

Kingdom’s administration has entirely decentralized health system governance to the island

governments of the constituent countries in line with their persistent demands of complete

autonomy over internal affairs. Besides matters of military defense and foreign policy, Dutch

administration has therefore been primarily concerned with the achievement of sustainable public

financial management in Aruba, Curaçao and St. Maarten in the past decades. This led to the

installation of the board of financial supervision (College Financieel Toezicht, CFT) in the islands

of the former Netherlands Antilles in 2010, and the board of financial supervision Aruba (CAft)

in Aruba in 2015. Collaborative programs in the area of health have mainly focused on aspects of

curative care, such as the organization of clinical conferences by the Nashko and the exchange of

medical students from Dutch teaching hospitals to the islands. In other cases, Dutch health

organizations and consultants have supported projects in the Dutch Caribbean, but these

collaborations were typically short-term. In May 2018, however, the four Ministers of Health from

the constituent countries of the Kingdom have committed to strengthening their cooperation,

among other things, in the areas of infectious diseases, emergency preparation, preventive and

curative care, health workforce and health information. In the special municipalities of the

Netherlands, Bonaire, Saba and St. Eustatius, governance is steered by the Dutch government

since the dissolution of the Netherlands Antilles on 10 October 2010.

Our results indicate that governments in Caribbean states where (health) organizations

collaboratively work with their counterparts in the country of their Western administrator are more

successful in improving and protecting the health of their population, likely because they have

more knowledge, tools and finance to their disposal, and have more effective bureaucracies to turn

plans into practice. Conversely, this also indicates that the halfhearted political situation with the

Netherlands has contributed to the current health situation in the Dutch Caribbean. In terms of

understanding the ambivalent governing style of the Dutch government towards its Dutch

Caribbean dependencies, it is important to place the current situation in its historical context.

Chapter 7

148

The accounts of the Dutch civil servant van Kol, who visited the Dutch Caribbean islands in the

first decades of the 20th century, suggest that the ambivalent governing style of the Dutch

government is not a contemporary phenomenon, but already characterized the Dutch colonial

administration of the islands at the time. In 1901, van Kol urgently pleaded the Dutch government

to prevent a further decline of living standards in the Dutch colony after years of economic

recession (65). In 1919, van Kol concluded that in spite of the progress made, the colonial

administrations on the islands had, so far, “ not fulfilled their duty with regard to public health”

(66). Over the years, when the settlement of oil refineries in Aruba and Curaçao had brought

economic growth, the colonial administration became more concerned with the social and

economic causes of ill health, as is illustrated by the reports by the Dutch civil servants Verwey

and de Gaay Fortman (67,68). As a result of the decolonization efforts, the 1954 Charter (Statuut)

for the Kingdom of the Netherlands now defined the relationship between the Netherlands and

the then newly established Netherlands Antilles as a voluntary and equal endeavor, in which mutual

assistance and cooperation is possible on request, and in which each country is largely autonomous

in arranging their own internal affairs (69). While all constituent countries of the Kingdom are

obliged to promote the realization of legal certainty, human rights and good governance conform

the Charter, the Kingdom council is the ultimate guarantor in all territories of the Kingdom. It did

not take long to realize that the Kingdom Council, however, has few means to intervene in the

island’s internal affairs and to hold the constituent’s countries’ governments responsible for their

(in)actions, yet is obliged to intervene in case of (financial) emergency on the request of the island’s

governments. Repeatedly, however, the island’s governments demonstrated that sustaining the

degree of autonomy over internal affairs took precedence over strengthening governmental

institutes, and hence, improving public services, to the level of the Kingdom (70).

Since the establishment of the Netherlands Antilles, the self-governing governments in the Dutch

Caribbean have not succeeded in closing the economic, educational and health gap with the

Netherlands, nor have they, as the results in this thesis suggest, addressed their population’s health

needs as effectively as other affiliated Caribbean states. On the contrary, the study in chapter 5 in

this thesis indicates that where life expectancy is concerned, the gap with the Netherlands has even

diverged in the past decades, in contrast to the more favorable population health outcomes in

other politically affiliated counterparts. The better health outcomes in Caribbean states with

stronger collaborations with their Western administrator suggests that Dutch Caribbean politicians

need to critically consider what degree of autonomy over internal affairs would serve the

population’s interest the most. In this light, it should also be noted that succeeding Dutch

governments, in turn, have been wittingly or unwittingly involved in endorsing policies that

General discussion

149

sustained the asymmetrical power relationships in postcolonial society (71), and that resulted in

weakened governmental bureaucracies through corporatization efforts since the 1980s (72).

Although the current political impasse may imply that discussions on the constitutional nature of

the Kingdom, and the scope of the authority of the Kingdom Council, are needed, this thesis

mostly calls for an increased awareness and determination to address the current health situation

in the Dutch Caribbean. In light of the recurring tensions around the shared colonial history

between the constituent countries of the Kingdom, we recognize that recommendations of

increased “Dutch influence” in the islands’ internal affairs to strengthen governmental

bureaucracies and to improve public services may be considered unwelcome by some. Although it

is apparent that these ideological dynamics influence the current discussions on the islands’

relationships with the Netherlands, a detailed discussion on this topic is beyond the scope of this

thesis on public health. It seems important to note, however, that the paradigm of neocolonialism,

a scholarly approach that criticizes Western influences in former colonies as this would increase

the gaps between rich and poor countries even further (73–76), predicts the exact opposite of what

is currently observed in the Caribbean region. Despite the negative connotation of the term

“(neo)colonialism”, this suggests that the judgement of contemporary relationships between the

Netherlands and the constituent countries in the Dutch Caribbean needs a more careful

consideration, for example based on how these relationships can contribute to the much needed

social and economic development in the former Dutch colonies. Under those circumstances, it

seems more likely that an agreement on all sides of the political spectrum can be reached, for

example on the undesirablity of further declines in population health on the Dutch Caribbean

islands relative to the Netherlands and to other politically affiliated Caribbean states.

Implications for research and policy

Based on the findings mentioned above, we now turn to a number of suggestions for future

research and policy efforts in the Dutch Caribbean. As special municipalities of the Netherlands,

health system governance on the islands of the Caribbean Netherlands (Bonaire, Saba and St.

Eustatius) is currently implemented with oversight from the Dutch government. In the absence of

comparative data, these territories were excluded from our studies. For the same reason, we were

unable to include St. Maarten. We will therefore focus our recommendations on Aruba and

Curaçao. The constituent countries of the Kingdom of the Netherlands govern their own health

sectors and steer their own health policy initiatives. We primarily recommend to invest in filling

Chapter 7

150

the gaps in the monitoring of population health and health services and in expanding the focus of

policy initiatives to areas of suboptimal performance.

If our findings on the better population health in Caribbean states that are politically affiliated to

France, the UK and the US provide any inspiration, it is that efforts to improve population health

in Aruba and Curaçao benefit from strong partnerships with (health) organizations in the

Netherlands. The Dutch Caribbean islands have relatively small populations, but the scope and

complexity of efforts needed for the adequate monitoring of population health and the effective

implementation of policy interventions require the efforts analogous to large countries. For

example, in the research efforts that are needed to regularly analyze data for a comprehensive set

of national health indicators from a variety of sources, such as health surveys and population-based

registers, the size of the population of interest is of marginal importance once data collection

measures are routinely implemented. Moreover, the extend of work required for the development,

implementation and evaluation of effective policy interventions for each specific policy area, for

instance the implementation of legislation and protocols to ensure quality standards of healthcare

services, is vast regardless of the size of the target population. Thus, the successful implementation

of the following recommendations will become more feasible in collaboration with larger (health)

organizations that are experienced in (health) research and policy implementation.

Implications for research in the Dutch Caribbean

Besides restricting our description of the health situation in the Dutch Caribbean, the gaps in

national health data availability in the Dutch Caribbean islands also illustrate that its current

application to meet the needs for relevant and timely information of local decision-makers (e.g.

politicians, physicians, health professionals, public health officers, and policy makers) to point out

issues, set priorities, support practices and monitor progress to reach certain targets in the Dutch

Caribbean is limited. More efforts in the monitoring of population health and healthcare services

in the Dutch Caribbean are therefore needed.

We recommend to establish a platform for health researchers in the Dutch Caribbean with as

primary aim to publish four-yearly population health reports that support the policy making

process at the strategic level. A shared health data platform was previously proposed during

consultations between the four Ministries of Health from the constituent countries of the

Kingdom in May 2018 (vierlandenoverleg), but thus far has not materialized. The population

health reports provide a broad analysis of the situation per Caribbean territory (i.e. health status,

risk factors, provision and quality of healthcare, care expenditures), and combine the available

General discussion

151

scientific evidence with background information (77). In order to achieve the regular publication

of these reports, we suggest the following steps:

First, the members of the platform should establish a health research strategy for the Dutch

Caribbean. This research strategy strives to optimize health outcomes by questioning what we can

do to improve population health (33) and adheres to the Essential National Health Research

(ENHR) principles (78). With more concentrated efforts, the Dutch Caribbean territories will be

in a better position to fulfill the research-related functions 1 and 11 of the Essential Public Health

Functions (EPHF), a set of actions formulated by the PAHO/WHO that contributes to the

broader context of health system strengthening (79). The health research strategy aims to establish

a sound and effective health information system in each territory, with a focus on the

harmonization and alignment of current research efforts and the identification and resolvement of

current short-comings in health research capacity and infrastructure. The aim of this is to establish

a baseline on which future progress can be monitored. A tangible outcome is a list of national

indicators, specifying definitions, periodicity of data collection and (potential) data sources, and

attuned to provide data for international initiatives such as the Sustainable Development Goals

(SDG’s) from the United Nations and health indicators from the Pan American Health

Organization/World Health Organization (PAHO/WHO).

Second, the members of the platform, or their related institutes, provide or guide technical support

for the collection, analysis and dissemination of data that are needed for the national indicators.

This will not only facilitate the production of comparative data across Dutch Caribbean territories,

but also optimize the available research capacity by resource sharing. For example, a researcher

specialized in mortality statistics can manage data collection and analysis for all territories. In cases

where a specific specialty is unavailable on the islands, the platform can facilitate collaborations

with Dutch health research institutes. To make these efforts sustainable, an approach is required

that makes health data collection, analysis and dissemination eventually a normal and routine part

of the policy-making process in each territory.

Three, the platform offers opportunities to stimulate the use of health information for evidence-

based policy practices in the Dutch Caribbean. For example, the members of the platform can

support interdisciplinary discussions on pressing health problems and their potential solutions,

organized for and with local stakeholders and decision-makers. This inquiry will also provide the

background information needed for the interpretation of the scientific evidence and

recommendations in the population health reports.

Chapter 7

152

Implications for policy in the Dutch Caribbean

This thesis showed that there is an urgent need for optimization of health services (chapter 3) and

implementation of health policies (chapter 6) in the Dutch Caribbean. The potential avoidability

of the specific causes-of-death, however, also suggests that in current health policy efforts, limited

focus is directed towards addressing the health needs of the Dutch Caribbean population. Setting

priorities for health policy should direct the efforts towards fixing this imbalance. We recommend

that this includes health policy areas related to the avoidable causes-of-death for which the current

mortality rates are disproportionally higher in the Dutch Caribbean than in the Netherlands and

the French islands (figure 1). The reason for this is that, in the absence of more detailed

information on morbidity, effective interventions in these areas are likely to produce the largest

health gains among the population.

It is important to note that the conclusions in this thesis are based on comparative analyses of

national outcomes, life expectancy data and cause-specific mortality rates in particular. The benefit

of national outcomes is that it allows for cross-territory comparisons to benchmark local health

outcomes. Nonetheless, these results alone do not give insight in what needs to be done for health

policy areas of suboptimal performance. In order to do this, the national outcomes need to be

linked with health system processes during consultations with stakeholders. For example, our study

in chapter 3 identified perinatal conditions as an important contributor to excess avoidable

mortality in Curaçao. For the next step, it is key to investigate access and quality of mother and

child care services in more detail so that certain aspects that are viable for effective targeting can

be identified. This assessment should determine what the shortcomings of current processes are,

and should identify the interventions and resources needed for improvements. Most likely,

incremental policy implementation will not be enough to improve national outcomes, and a

number of complementary interventions will be selected. Strategies that are recommended by the

PAHO/WHO were previously proven effective and can serve as guidelines to achieve effective

policy making.

As Oliver notes, however, “science can identify solutions to pressing public health problems, but

only politics can turn most of those solutions into reality” (80). Past experiences also reveal that it

would be naïve to assume that political commitment automatically translates in tangible results

(81). Thus, resource allocation, both financial and human capital, need to reflect the commitment

to improving population health. By focusing on specific health policy areas of suboptimal

performance, these concentrated policy efforts are likely to stimulate the much needed health

progress in the Dutch Caribbean.

General discussion

153

Overall conclusion

Overall, this thesis gives a better insight into the health situation in the Dutch Caribbean, and the

factors related to this health situation, the political context and health policy performance in

particular. We show that the health status of the Dutch Caribbean population in Aruba and

Curaçao is poorer than in the Netherlands and in other politically affiliated states (chapter 2, 3

and 5). People in Aruba and Curaçao were more likely to die from causes that are avoidable in the

presence of timely and effective healthcare and/or interventions in public health and prevention

than in the Netherlands, Guadeloupe and Martinique: breast cancer, cervix cancer, ischemic heart

disease, cerebrovascular disease, pneumonia, nephrosis/nephritis and transport accidents. In

Curaçao, important additional contributors of excess mortality are HIV/AIDS, cancer of the colon

and rectum, diabetes, hypertensive heart disease, perinatal deaths and violence. This suggests that

an important aspect contributing to the poorer mortality outcomes in the Dutch Caribbean is that

the local governments have, so far, not optimally addressed their population’s health needs.

Therefore, health outcomes in the Dutch Caribbean can potentially improve with targeted actions

that improve the access and quality of health care services (chapter 3) and stimulate the

implementation of “best practice“ health policies (chapter 6).

Like other studies before us, we demonstrated a positive effect of a strong political affiliation to a

Western country on population health and, relatedly, health policy performance, in Caribbean

states (chapter 4, 5 and 6). The studies in this thesis expand the existing knowledge in two

important ways. First, we show that life expectancy in currently sovereign states was already

unfavorable than in currently affiliated states during the colonial period, that decolonization

coincided with a reduction in life expectancy growth, and that the gap in life expectancy between

current sovereign and affiliated states continues to increase to the present day (chapter 4). Second,

we show that the differences in life expectancy trends of Caribbean states that are politically

affiliated to France, the UK and the US, and the Netherlands reflect the ways in which these states

are governed. The trends in life expectancy are more favorable in territories with less political

autonomy, Martinique, Guadeloupe, and the British Islands, than in territories with more political

autonomy, Puerto Rico, the US Virgin Islands, Aruba and Curaçao (chapter 5). The degree of

political autonomy translates into various governance structures in the area of public health, which

suggests that local governments in Caribbean states where health organizations collaboratively

work with their corresponding organizations in the country of their Western administrator are

more successful in improving and protecting of the health of their population. For example,

Martinique and Guadeloupe were more likely to have implemented “best practice” health policies

Chapter 7

154

(chapter 6), possibly because territories with less political autonomy have more human capital,

financial resources, investments and technologies at their disposal to identify and address health

problems, but also because they have more effective bureaucracies to turn plans into practice.

Conversely, this also indicates that the halfhearted political situation between the Dutch Caribbean

islands and the Netherlands, in which discussions about the political autonomy and financial

sustainability of the islands have dominated in recent decades, contributed to the current health

situation in the Dutch Caribbean. Although this may imply that discussions about the

constitutional character of the Kingdom of the Netherlands and the scope of responsibility of the

Kingdom council are necessary, this mostly calls for an increased awareness of the increasing

health deprivation of the Dutch Caribbean population relative to the populations in the

Netherlands and in other politically affiliated Caribbean states, as well as a determination to address

the current health situation.

Future health research on the Dutch-Caribbean islands should focus on the monitoring of

population health and healthcare so that relevant and timely information can be provided to

decision-makers. To improve population health in the Dutch Caribbean, we recommend

prioritizing health policy areas of suboptimal performance, as effective interventions in these areas

are likely to produce the largest health gains among the population. If our findings on the more

favorable health situation in Martinique, Guadeloupe and the British Islands offer any inspiration,

it is that the feasibility of the successful implementation of these recommendations benefits from

strong collaborations with Dutch (health) organizations.

General discussion

155

References

1. Stoeldraijer L. Bevolkingstrends 2014. Levensverwachting in Caribisch Nederland verschilt weinig met Nederland. Centraal Bureau voor de Statistiek; 2014. 1–10 p.

2. Statistics Netherlands. Leefstijl, gezondheid, zorggebruik Caribisch Nederland [Internet]. 3 June 2019. 2019. Available from: https://www.cbs.nl/nl-nl/maatwerk/2019/23/leefstijl-gezondheid-zorggebruik-caribisch-nederland


4. Boyko EJ. Observational research- opportunities and limitations. J Diabetes Complications. 2013;27(6):642–8.

5. Black N. Why we need observational studies to evaluate the effectiveness of health care. BMJ. 1996;312(7040):1215–8.

6. Wang MTM, Bolland MJ, Grey A. Reporting of limitations of observational research. JAMA Intern Med. 2015;175(9):1571–2.

7. Galea S, Riddle M, Kaplan GA. Causal thinking and complex system approaches in epidemiology. Int J Epidemiol. 2010;39(1):97–106.




11. Hill AB. The environment and disease: association or causation? Proc R Soc Med. 1965;58(5):295–300.




15. Caldwell JC. Routes to Low Mortality in Poor Countries. Popul Dev Rev. 1986;12(2):171–220.

16. Mackenbach JP. Cultural values and population health: a quantitative analysis of variations in cultural values, health behaviours and health outcomes among 42 European countries. Heal Place. 2014;28:116–32.

17. Feyrer J, Sacerdote B. Colonialism and Modern Income – Islands as Natural Experiments. Cambridge; 2006. (NBER Working Paper Series). Report No.: 12546.

18. Mackenbach JP, Mckee M, editors. Successes and Failures of Health Policy in Europe. Four decades of divergent trends and converging challenges. Buckingham: Open University Press; 2013.

19. Mackenbach JP, McKee M. A comparative analysis of health policy performance in 43 European

Chapter 7

156

countries. Eur J Public Health. 2013;23(2):195–201.

20. Office for National Statistics. Definition of avoidable mortality. Newport: ONS; 2012. 1–7 p.

21. Vierbergen F. Inkomens en inkomensverdelingen in Curaçao - publicatiereeks Census 2011. Willemstad; 2013.

22. Verstraeten S. The Curaçao Global School-based Student Health Survey (GSHS). Willemstad: Volksgezondheid Instituut Curaçao; 2015.

23. Steil J, Menendian S, Gambhir S. Responding to rising inequality. Policy interventions to ensure opportunity for all. Berkeley: Haas Institute for a Fair and Inclusive Society; 2014.

24. Humpel N, Owen N, Leslie E. Environmental factors associated with adults’ participation in physical activity: a review. Am J Prev Med. 2002;22(3):188–99.

25. Saelens BE, Handy SL. Built environment correlates of walking: A review. Med Sci Sports Exerc. 2008;40(7 Suppl):S550-566.

26. Fleischhacker SE, Evenson KR, Rodriguez DA, Ammerman AS. A systematic review of fast food access studies. Obes Rev. 2011;12(5):e460-471.

27. Pulster EL, Johnson G, Hollander D, McCluskey J, Harbison R. Exposure Assessment of Ambient Sulfur Dioxide Downwind of an Oil Refinery in Curaçao. J Environ Prot (Irvine, Calif). 2018;9(3):194–210.

28. van der Torn P. Health complaints & air pollution from the Isla refinery in Curaçao; with special emphasis to the response to irregular situations. Rotterdam: Public Health Service Rotterdam; 1999.

29. Pin R, Verstraeten S, Griffith-Lendering M. Nationale Gezondheidsenquête Curaçao. Themarapport gezondheid, leefstijl en zorggebruik in de arme wijken. Willemstad: Volksgezondheid Instituut Curaçao; 2017.

30. Alkire BC, Peters AW, Shrime MG, Meara JG. The economic consequences of mortality amenable to high-quality health care in low- and middle-income countries. Health Aff. 2018;37(6):988–96.

31. Kruk ME, Gage AD, Joseph NT, Danaei G, García-Saisó S, Salomon JA. Mortality due to low-quality health systems in the universal health coverage era: a systematic analysis of amenable deaths in 137 countries. Lancet. 2018;392(10160):2203–12.

32. Central Bureau of Statistics Curaçao. Statistical by subject: drop-outs and youth unemployment rates [Internet]. 2019 [cited 2019 Aug 5]. Available from: www.cbs.cw

33. Grillet ME, Hernández-Villena J V, Llewellyn MS, Paniz-Mondolfi AE, Tami A, Vincenti-Gonzalez MF, et al. Venezuela’s humanitarian crisis, resurgence of vector-borne diseases, and implications for spillover in the region. Lancet Infect Dis. 2019;3099(18):1–13.

34. Paniz-Mondolfi AE, Tami A, Grillet ME, Márquez M, Hernández-Villena J, Escalona-Rodríguez MA, et al. Resurgence of vaccine-preventable diseases in Venezuela as a regional public health threat in the Americas. Emerg Infect Dis. 2019;25(4):625–32.

35. Zavras D, Tsiantou V, Pavi E, Mylona K, Kyriopoulos J. Impact of economic crisis and other demographic and socio-economic factors on self-rated health in Greece. Eur J Public Health. 2013;23(2):206–10.

36. Karanikolos M, Mladovsky P, Cylus J, Thomson S, Basu S, Stuckler D, et al. Financial crisis, austerity, and health in Europe. Lancet. 2013;381(9874):1323–31.

37. Dagevos J, Dagevos M. Minderheden meer gewicht. Over overgewicht bij Turken, Marokkanen,

General discussion

157

Surinamers en Antillianen en het belang van integratiefactoren. Den Haag: Sociaal en Cultureel Planbureau; 2008.

38. Lindeman E, Booi H, Cohen L. Monografie Antillianen 2010. De positie van Antillianen en Arubanen in Amsterdam. Amsterdam: Gemeente Amsterdam; 2010.

39. Schouten G, Stam B, Christiaanse B, Van Veelen J. Gezondheidsenquête 2008. Onderzoeksmethode en gezondheid in relatie tot achtergrondkenmerken. Rotterdam: GGD Rotterdam-Rijnmond; 2010.

40. Stirbu I, Kunst a E, Bos V, Mackenbach JP. Differences in avoidable mortality between migrants and the native Dutch in The Netherlands. BMC Public Health. 2006;6(78).

41. Rijksinsituut voor de Volksgezondheid en Milieu. RIVM rapport 441100019. HIV-surveys bij hoog-risicogroepen in Rotterdam 2002-2003. Bilthoven: RIVM; 2005.

42. Proctor A, Krumeich A, Meershoek A. Making a difference: The construction of ethnicity in HIV and STI epidemiological research by the Dutch National Institute for Public Health and the Environment (RIVM). Soc Sci Med. 2011;72(11):1838–45.

43. Vandenheede H, Deboosere P, Stirbu I, Agyemang CO, Harding S, Juel K, et al. Migrant mortality from diabetes mellitus across Europe: The importance of socio-economic change. Eur J Epidemiol. 2012;27(2):109–17.

44. ISOG. Project obesitas preventie op Curaçao. Deelproject 1. Willemstad: Eilandgebied Curaçao; 2008.

45. Roux AVD. The study of group-level factors in epidemiology: Rethinking variables, study designs, and analytical approaches. Epidemiol Rev. 2004;26:104–11.

46. Berkman LF. Introduction: seeing the forest and the trees - from observation to experiments in social epidemiology. Epidemiol Rev. 2004;26(1):2–6.

47. Macintyre S, Ellaway A, Cummins S. Place effects on health: how can we conceptualise, operationalise and measure them? Soc Sci Med. 2002;55(1):125–39.

48. Rose G. Sick Individuals and Sick Populations. Int J Epidemiol. 1985;30(3):427–32.

49. Central Bureau of Statistics Aruba. Fourth population and housing census Aruba. Oranjestad; 2000.

50. ter Bals M. Demography of Curaçao, Publication Series Census 2011. Willemstad: CBS Curaçao; 2014.

51. Taras V, Steel P, Kirkman BL. Does country equate with culture? Beyond geography in the search for cultural boundaries. Manag Int Rev. 2016;56(4):455–87.

52. Algemene Rekenkamer Curaçao. Chikungunya, aanpak van de bestrijding. Willemstad: Algemene Rekenkamer Curaçao; 2016.

53. Roe AE. The sound of silence: ideology of national identity and racial inequality in contemporary Curaçao. Miami: Florida International University; 2016.

54. Huijnk W, Andriessen I. Integratie in zicht? De integratie van migranten in Nederland op acht terreinen nader bekeken. Den Haag: Sociaal en Cultureel Planbureau; 2016.

55. De Juan A, Pierskalla JH. The comparative politics of colonialism and its legacies: an introduction. Polit Soc. 2017;45(2):159–72.

56. Woodberry RD. The missionary roots of liberal democracy. Am Polit Sci Rev. 2012;106(2):244–74.

Chapter 7

158

57. Acemoglu D, Johnson S, Robinson JA. The colonial origins of comparative development: an empirical investigation. American Economic Review. Cambridge; 2000. (NBER; vol. 91). Report No.: 7771.


59. Gilley B. The case for colonialism. Acad Quest. 2018;31(2):167–85.

60. McElroy JL, Mahoney M. The propensity for political dependence in island microstates. Insul Int J Isl Aff. 1999;9:32–5.


62. Acemoglu D, Robinson J. Why Nations Fail: The Origins of Power, Prosperity, and Poverty. 1st ed. New York: Crown; 2012.

63. Pan American Health Organization (PAHO/WHO). Health in the Americas. Country volume: French Guiana, Guadeloupe, and Martinique. Washington: PAHO/WHO; 2012.

64. Department for International Development. Health and healthcare in the British Overseas Territories: Regional and UK government support. 2010.

65. van Kol H. Een noodlijdende kolonie. Princenhage: Masereeuw & Bouten; 1901.

66. van Kol H. De volksgezondheid in onze West-Indische Koloniën. Nieuwe West Indische Gids. 1919;1(1).

67. Verwey RA. De sociale ontwikkelingsgang van Curaçao. West-Indische Gids. 1935;161–74.

68. de Gaay Fortman B. Economische en sociale vraagstukken in Curaçao. West-Indische Gids. 1938;329–41.

69. Charter for the Kingdom of the Netherlands. Ned Tijdschr Voor Int R. 1958;5(1):107–18.

70. Oostindie G. Dependence and autonomy in sub-national island jurisdictions: the case of the Kingdom of the Netherlands. Round Table Commonw J Int Aff. 2006;(386):609–26.

71. Groenewoud M. “Nou koest, nou kalm”. De ontwikkeling van de Curaçaose samenleving, 1915-1973. Willemstad: Universiteit Leiden; 2017.

72. Goede M. Groei en fragmentatie van de overheid op de Nederlandse Antillen en Curaçao. Willemstad: Universiteit van Tilburg; 2005.

73. Cardoso FH, Faletto E. Dependency and Development in Latin America. 1st ed. University of California Press; 1979.


75. Chiriyankandath J. Colonialism and Post-Colonial Development. In: Burnell P, Randall V, Rakner L, editors. Politics in the Developing World. Third. New York: Oxford University Press; 2011.

76. Girvan N. Colonialism and neo-colonialism in the Caribbean: an overview. In: IV International Seminar Africa, The Caribbean and Latin America. St. Vincent and the Grenadines: University of the West Indies; 2012.

77. Verschuuren M, van Oers H, editors. Population Health Monitoring. Climbing the Information Pyramid. 1st ed. Cham: Springer International Publishing; 2019.

General discussion

159

78. Task Force on Health Research for Development Secretariat. Essential National Health Research. A Strategy for Action in Health and Human Development. Geneve: United Nations; 1991.

79. Pan American Health Organization (PAHO/WHO). The essential public health functions as a strategy for improving overall health systems performance: trends and challenges since the public health in the Americas initiative, 2000-2007. 2008.

80. Oliver TR. The politics of public health policy. Annu Rev Public Health. 2006;27(1):195–233.


English summary

English summary

162

The Netherlands is at the forefront of population health monitoring, but not much is known about

the health of inhabitants in the Caribbean territories of the Kingdom of the Netherlands. The

information that is available seems to fit Sir Michael Marmot’s statement that “all too commonly

where health is poorest, health information tends to be poorest”. Life expectancies are shorter,

and infant and maternal mortality rates are higher, in the two largest Dutch Caribbean islands,

Aruba and Curaçao, than in the Netherlands. The causes of the poorer health outcomes of the

Dutch Caribbean people are largely unknown, but are commonly ascribed to unspecified regional,

biological or cultural characteristics. Beyond these general descriptions, however, more specific

explanations are necessary to inform policy strategies that are aimed at improving the health of

inhabitants in the Dutch Caribbean islands.

The main objective of this thesis is to provide a better insight into the health situation of the Dutch

Caribbean, and the factors related to this health situation, in particular the role of the political

context and health policy performance. In order to address this aim, we have made use of data

that are derived from cross-sectional health surveys, mortality registration systems, harmonized

international databases, and country reports. To address the multidisciplinary aspects of the

determinants of population health, we cover theory from the fields of public health, medicine,

political science, organization science, economics, and sociology.

First, we sought to explore how the health status of the Dutch Caribbean population compares to

the Dutch population, and to the populations of other Caribbean states. We considered three

different health measures: self-reported health indicators, life expectancy and cause-specific

mortality rates. Overall, we found that the health status of the Dutch Caribbean population is

poorer than that of the Dutch population (chapter 2 and 3), and that of populations in Caribbean

states politically affiliated to France, the United Kingdom (UK), and the United States (US) as well

(chapter 5).

In chapter 2, we made a three-way comparison between the Dutch Caribbean population living

in Curaçao, Dutch Caribbean migrants living in the Netherlands, and non-migrants living in the

Netherlands. The results of chapter 2 indicate that, perhaps surprisingly, the Dutch Caribbean

population reported better physical and mental health than Dutch Caribbean migrants in the

Netherlands. Health outcomes were poorer for Dutch Caribbean migrants compared to non-

migrants living in the Netherlands as well. These results persisted even when controlling for

sociodemographics (age, marital status, education and employment), risk factors (tobacco and

alcohol use, overweight and physical activity) and urbanization (in the Netherlands). The

exceptions were overweight/obesity and diabetes; prevalence rates for overweight/obesity (men

English summary

163

only) and diabetes (both genders) were similar for the Dutch Caribbean populations living in

Curaçao and in the Netherlands, and higher than in Dutch non-migrants. For Aruba, no

comparative data were available to include in this study.

The study in chapter 3 demonstrates that levels of amenable mortality, deaths from causes that

are considered unnecessary in the presence of timely and effective healthcare, have been

consistently higher in Aruba and the NA/Curaçao than in the Netherlands during the 1988-2014

period. If mortality from amenable causes were reduced to similar levels as in the Netherlands,

men and women in Aruba would add respectively 1.19 years and 0.72 years to their life

expectancies. In Curaçao, this would be 2.06 years and 2.33 years. If mortality from amenable

circulatory causes (which includes ischemic heart disease, cerebrovascular disease, and

hypertensive disease) were reduced to similar levels as in the Netherlands, men and women in

Aruba would add respectively 1.02 years and 0.56 years to their life expectancies. In NA/Curaçao,

this would be 0.65 years and 0.55 years. Perinatal deaths are an important additional contributing

cause-of-death in NA/Curaçao, and are responsible for 0.79 years (men) and 0.90 years (women)

of the total mortality gap with the Netherlands. Nephritis/nephrosis is also important for

NA/Curaçao, where it contributes 0.12 years for men and 0.21 years for women. In addition,

reduction in breast cancer mortality to the level of the Netherlands would add 0.22 years in Aruba

and 0.21 years in NA/Curaçao to female life expectancy.

The results of chapter 5 show that Aruba and NA/Curaçao experienced declines in male and

stagnation in female life expectancy, and that life expectancy trends have diverged relative to the

Netherlands during the 1980-2014 period. In contrast, life expectancy trends in Caribbean

dependencies that had relatively low levels of autonomy, Martinique, Guadeloupe and the UK

islands, ran in parallel with those in their administrator countries. Decomposition of life expectancy

differences between Western countries and their dependencies point to the important contribution

of mortality from cardiovascular diseases and external causes to the excess mortality in the

Caribbean states. The magnitude of mortality differences with Western countries, however, differs

greatly. Unlike other affiliated states, Dutch dependencies did not experience substantial

improvements in cardiovascular mortality during the 1980-2014 period. Also, NA/Curaçao

experienced sharp rises in homicide during the study period, alongside US territories Puerto Rico

and the US Virgin Islands. In addition, transport accidents are an important additional contributor

to higher external cause mortality in NA/Curaçao, and the main contributor to higher external

cause mortality in Aruba.

English summary

164

Second, we assessed the role of the political context and health policy performance as potential

factors that influence population health variations between Caribbean states. We considered three

determinants for the political context: current political sovereignty status, decolonization and

degree of political affiliation. Overall, we found that decolonization has increased the life

expectancy gap between currently sovereign and affiliated Caribbean states, and that a strong

political affiliation to a Western country is positively related to population health and health policy

performance in Caribbean states.

In chapter 4, we assessed how disparities in life expectancy between currently sovereign and

affiliated Caribbean states have developed over time. We saw that life expectancies in currently

sovereign states were already lower than in currently affiliated states before the decolonization

events in the last half of the 20th century. Decolonization, the process in which a colony becomes

politically independent, coincided with reductions in life expectancy growth in the decade after

independence, which did not correspond with similar changes in economic performance. After the

decolonization events, the gap in life expectancy between currently sovereign and affiliated states

accelerated in the 1990's, and it continues to increase to the present day.

In chapter 5, we investigated whether differences in life expectancy trends between Caribbean

dependencies and their Western administrators were related to their degree of political

independence. We found that Caribbean states with a strong political affiliation to their Western

administrators, Martinique, Guadeloupe, and the UK islands, experienced more favorable life

expectancy trends during the 1980-2014 period than Caribbean states with higher levels of

autonomy, the US Virgin Islands, Puerto Rico, Aruba and Curaçao. In comparison with the

Western administrator, the underlying trends in causes-of-death are similar among Caribbean

states and possibly reflect region-specific drivers. The politically affiliated Caribbean states,

however, differ greatly in the magnitude of changes in mortality, most notably for cardiovascular

diseases and external causes. The least favorable trends are found for affiliated states with a high

degree of political autonomy.

In chapter 6 of this thesis, we determined whether and why 16 Caribbean states vary in health

policy performance in 11 different policy areas related to regional health objectives of the Pan

American Health Organization (PAHO) during the 2010-2015 period: 1. HIV/AIDS, 2.

Communicable disease, 3. Cancer screening, 4. Tobacco, 5. Fertility, pregnancy and childbirth, 6.

Child health, 7. Diabetes, 8. Hypertension, 9. Alcohol, 10. Road safety, and 11. Violence. Our

findings suggest that variations in population health outcomes between Caribbean states are related

to differences in the implementation of effective health policies. In other words, Caribbean states

English summary

165

with better health outcomes were more likely to have implemented a comprehensive set of best

practice health policies for the 11 policy areas included in this study. Martinique, Cuba and

Guadeloupe had the highest health policy performance summary scores, and Guyana, Belize and

Suriname had the lowest. Aruba has the lowest health policy performance summary score of five

politically affiliated Caribbean states included in this study, indicating that performance in 11 policy

areas related to important regional health objectives is suboptimal compared to other politically

affiliated Caribbean states. In terms of health outcomes, Aruba performs considerably worse on

breast cancer, lung cancer and road traffic mortality, and on teenage pregnancies. The latest

available mortality data from Curaçao dated from 2007 and we therefore had to refrain from doing

a similar analysis for Curaçao.

Again, we found a strong association with political sovereignty status: Caribbean states that

remained politically affiliated to Western countries had generally higher health policy performance

scores. Health policy performance was also better in states where the population density, GDP

per capita, and control of corruption is high, and states that are less religiously fragmented. In this

analysis, Cuba stands out as an overachiever considering its political independence, its low

economic development and its exceptional health outcomes. The potential for mortality reduction

from causes that are avoidable by health policy interventions varies greatly between Caribbean

states, and is considerable for the Caribbean as a whole: if the mortality rates of Martinique applied

in other Caribbean states, an overall mortality reduction of 12% could be obtained.

Overall, this thesis gives a better insight into the health situation in the Dutch Caribbean, and the

factors related to this health situation, the political context and health policy performance in

particular. We show that the health status of the Dutch Caribbean population in Aruba and

Curaçao is poorer than in the Netherlands and in other politically affiliated states (chapter 2, 3

and 5). People in Aruba and Curaçao were more likely to die from causes that are avoidable in the

presence of timely and effective healthcare and/or interventions in public health and prevention

than in the Netherlands, Guadeloupe and Martinique: breast cancer, cervix cancer, ischemic heart

disease, cerebrovascular disease, pneumonia, nephrosis/nephritis and transport accidents. In

Curaçao, important additional contributors of excess mortality are HIV/AIDS, cancer of the colon

and rectum, diabetes, hypertensive heart disease, perinatal deaths and violence. This suggests that

an important aspect contributing to the poorer mortality outcomes in the Dutch Caribbean is that

the local governments have, so far, not optimally addressed their population’s health needs.

Therefore, health outcomes in the Dutch Caribbean can potentially improve with targeted actions

English summary

166

that improve the access and quality of health care services (chapter 3) and stimulate the

implementation of “best practice“ health policies (chapter 6).

Like other studies before us, we demonstrated a positive effect of a strong political affiliation to a

Western country on population health and, relatedly, health policy performance, in Caribbean

states (chapter 4, 5 and 6). The studies in this thesis expand the existing knowledge in two

important ways. First, we show that life expectancy in currently sovereign states was already

unfavorable than in currently affiliated states during the colonial period, that decolonization

coincided with a reduction in life expectancy growth, and that the gap in life expectancy between

current sovereign and affiliated states continues to increase to the present day (chapter 4). Second,

we show that the differences in life expectancy trends of Caribbean that are politically affiliated to

France, the UK and the US, and the Netherlands reflect the ways in which these states are

governed. The trends in life expectancy are more favorable in territories with less political

autonomy, Martinique, Guadeloupe, and the British Islands, than in territories with more political

autonomy, Puerto Rico, the US Virgin Islands, Aruba and Curaçao (chapter 5). The degree of

political autonomy translates into various governance structures in the area of public health, which

suggests that local governments in Caribbean states where health organizations collaboratively

work with their corresponding organizations in the country of their Western administrator are

more successful in improving and protecting of the health of their population. For example,

Martinique and Guadeloupe were more likely to have implemented “best practice” health policies

(chapter 6), possibly because territories with less political autonomy have more human capital,

financial resources, investments and technologies at their disposal to identify and address health

problems, but also because they have more effective bureaucracies to turn plans into practice.

Conversely, this also indicates that the halfhearted political situation between the Dutch Caribbean

islands and the Netherlands, in which discussions about the political autonomy and financial

sustainability of the islands have dominated in recent decades, contributed to the current health

situation in the Dutch Caribbean. Although this may imply that discussions about the

constitutional character of the Kingdom of the Netherlands and the scope of responsibility of the

Kingdom council are necessary, this mostly calls for an increased awareness of the increasing

health deprivation of the Dutch Caribbean population relative to the populations in the

Netherlands and in other politically affiliated Caribbean states, as well as a determination to address

the current health situation.

Future health research on the Dutch-Caribbean islands should focus on the monitoring of

population health and healthcare so that relevant and timely information can be provided to

English summary

167

decision-makers. To improve population health in the Dutch Caribbean, we recommend

prioritizing health policy areas of suboptimal performance, as effective interventions in these areas

are likely to produce the largest health gains among the population. If our findings on the more

favorable health situation in Martinique, Guadeloupe and the British Islands offer any inspiration,

it is that the feasibility of the successful implementation of these recommendations benefits from

strong collaborations with Dutch (health) organizations.

Nederlandse samenvatting


170

Nederland loopt voorop in het monitoren van de volksgezondheid, maar er is weinig bekend over

de gezondheid van de inwoners van de Caribische gebieden in het Koninkrijk der Nederlanden.

De informatie die beschikbaar is komt overeen met de uitspraak van Sir Michael Marmot dat "maar

al te vaak waar de gezondheid gering is, is de informatie over de gezondheid ook gering". De

levensverwachting is korter en de kinderen moedersterfte zijn hoger op de twee grootste

Nederlands-Caribische eilanden, Aruba en Curaçao, dan in Nederland. De oorzaken van de

ongunstigere gezondheidsuitkomsten onder de Nederlands-Caribische bevolking zijn grotendeels

onbekend, maar worden vaak toegeschreven aan -verder- niet-gespecificeerde regionale,

biologische of culturele kenmerken. Naast deze algemene omschrijvingen zijn echter meer gerichte

verklaringen nodig voor de ondersteuning van beleidsstrategieën voor het verbeteren van de

gezondheid van de inwoners op de Nederlands-Caribische eilanden.

Het doel van dit proefschrift is om een beter inzicht te geven in de gezondheidssituatie van de

Nederlands-Caribische bevolking en de factoren die verband houden met deze

gezondheidssituatie, in het bijzonder de rol van de politieke context en de prestaties van het

gezondheidsbeleid. Om dit doel te bereiken maakten we gebruik van data uit cross-sectionele

gezondheidsenquêtes, sterfteregistraties, geharmoniseerde internationale databases en rapporten.

Voor de beschouwing van de multidisciplinaire aspecten van gezondheidsdeterminanten

behandelen we literatuur uit de disciplines publieke gezondheid, geneeskunde, politieke

wetenschappen, organisatiekunde, economie en sociologie.

Op de eerste plaats onderzochten we hoe de gezondheidstoestand van de Nederlands-Caribische

bevolking zich verhoudt tot die van de Nederlandse bevolking en de bevolking in andere

Caribische gebieden. We bekeken drie verschillende gezondheidsmaten: zelfgerapporteerde

gezondheidsindicatoren, de levensverwachting en oorzaakspecifieke sterftecijfers. We laten zien

dat de gezondheidstoestand van de Nederlands-Caribische bevolking minder gunstig is dan die

van de Nederlandse bevolking (hoofdstuk 2 en 3) en ook dan die van de bevolking van Caribische

staten die politiek verbonden zijn met Frankrijk, het Verenigd Koninkrijk (VK) en de Verenigde

Staten (VS) (hoofdstuk 5).

In hoofdstuk 2 maakten we een drievoudige vergelijking tussen de Nederlands-Caribische

bevolking op Curaçao, de Nederlands-Caribische migranten in Nederland en de niet-migranten in

Nederland. De resultaten in hoofdstuk 2 geven aan dat, misschien verrassend, de Nederlands-

Caribische bevolking op Curaçao een betere lichamelijke en geestelijke gezondheid rapporteerde

dan de Nederlands-Caribische migranten in Nederland. De gezondheidsuitkomsten van


171

Nederlands-Caribische migranten is ook minder gunstig dan die van niet-migranten in Nederland.

Deze resultaten bleven significant wanneer we corrigeerden voor sociodemografie (leeftijd,

burgerlijke staat, opleiding en werk), risicofactoren (tabak- en alcoholgebruik, overgewicht en

lichamelijke activiteit) en verstedelijking (in Nederland). De uitzonderingen waren

overgewicht/obesitas en diabetes; de prevalenties voor overgewicht/obesitas (alleen mannen) en

diabetes (beide geslachten) onder de Nederlands-Caribische bevolking op Curaçao en in

Nederland waren vergelijkbaar, en hoger dan voor Nederlandse niet-migranten. Voor Aruba waren

geen vergelijkende gegevens beschikbaar om in deze studie op te nemen.

De resultaten in hoofdstuk 3 laten zien dat de sterfte aan vermijdbare oorzaken, oorzaken die in

de aanwezigheid van tijdige en effectieve gezondheidszorg als vermijdbaar worden beschouwd,

tussen 1988 en 2014 consistent hoger waren op Aruba en op de NA/Curaçao dan in Nederland.

Als de vermijdbare sterfte wordt teruggebracht tot een vergelijkbaar niveau als in Nederland,

zouden mannen en vrouwen op Aruba respectievelijk 1,19 jaar en 0,72 jaar toevoegen aan hun

levensverwachting. Op de NA/Curaçao zou dit 2,06 jaar en 2,33 jaar zijn. Als de sterfte door

vermijdbare oorzaken van de bloedsomloop (waaronder ischemische hartziekte, cerebrovasculaire

aandoeningen en hypertensieve aandoeningen) wordt verlaagd tot een vergelijkbaar niveau als in

Nederland, zouden mannen en vrouwen op Aruba respectievelijk 1,02 jaar en 0,56 jaar toevoegen

aan hun levensverwachting. Op de NA/Curaçao zou dit 0,65 jaar en 0,55 jaar zijn. Perinatale

sterfgevallen zijn een belangrijke additionele doodsoorzaak op de NA/Curaçao en zijn

verantwoordelijk voor 0,79 jaar (mannen) en 0,90 jaar (vrouwen) van de totale mortaliteitskloof

met Nederland. Nefritis/nefrose is ook belangrijk voor de NA/Curaçao en draagt 0,12 jaar bij

voor mannen en 0,21 jaar voor vrouwen. Een vermindering van de borstkankersterfte tot het

niveau van Nederland zou op Aruba 0,22 jaar en op NA/Curaçao 0,21 jaar aan de vrouwelijke

levensverwachting toevoegen.

De resultaten in hoofdstuk 5 geven aan dat Aruba en de NA/Curaçao tussen 1980 en 2014 een

daling van de mannelijke en een stagnatie van de vrouwelijke levensverwachting hebben ervaren.

De trends in levensverwachting zijn daarmee gedivergeerd ten opzichte van Nederland. De

levensverwachtingstrends in politiek afhankelijke Caribische staten met relatief weinig autonomie,

Martinique, Guadeloupe en de Britse eilanden, liepen daarentegen parallel met die van hun

westerse bestuurder. Een verdere uitsplitsing van de verschillen in levensverwachting tussen

westerse landen en hun geaffilieerde Caribische gebieden toont dat de sterfte aan hart- en

vaatziekten en externe oorzaken een belangrijke bijdrage levert aan de overtollige sterfte in de

Caribische gebieden. De omvang van sterfteverschillen met westerse landen verschilt echter sterk.


172

In tegenstelling tot andere geaffilieerde gebieden verbeterde de cardiovasculaire sterfte in de

Nederlands-Caribische staten tussen 1980 en 2014 niet substantieel. De NA/Curaçao maakte ook

een sterke stijging van het aantal moorden mee tijdens de onderzoeksperiode, evenals de

Amerikaanse gebieden Puerto Rico en de Amerikaanse Maagdeneilanden. Daarnaast draagt sterfte

aan verkeersongevallen ook in belangrijke mate bij aan het hoge sterftecijfer voor externe oorzaken

op de NA/Curaçao, en is dat de belangrijkste oorzaak van de hogere sterfte aan externe oorzaken

op Aruba.

Op de tweede plaats onderzochten we twee potentiële factoren die de gezondheidsverschillen

tussen Caribische staten beïnvloeden, de politieke context en de prestaties van het

gezondheidsbeleid. We bekeken drie factoren voor de politieke context: de huidige politieke status,

dekolonisatie en de mate van politieke autonomie. We laten zien dat dekolonisatie het verschil in

levensverwachting tussen de huidige soevereine en geaffilieerde Caribische staten heeft vergroot

en dat een sterke politieke band met een westers land positief gerelateerd is aan de

gezondheidsuitkomsten en de prestaties van het gezondheidsbeleid in Caribische staten.

In hoofdstuk 4 bekeken we hoe de verschillen in levensverwachting tussen de huidige soevereine

en geaffilieerde Caribische staten zich in de loop van de tijd hebben ontwikkeld. We zagen dat de

levensverwachting in de huidige soevereine staten al lager was dan in de huidige geaffilieerde staten

vóórdat dekolonisatie in de laatste helft van de 20e eeuw plaats vond. Dekolonisatie, het proces

waarbij een kolonie politiek onafhankelijk wordt, viel samen met een vermindering van de groei in

levensverwachting in het decennium na de onafhankelijkheid, maar niet met soortgelijke

veranderingen in economische groei. In de jaren 90 wijkt de kloof in levensverwachting tussen de

huidige soevereine en geaffilieerde staten versneld uit elkaar, en deze blijft tot op de dag van

vandaag toenemen.

In hoofdstuk 5 onderzochten we of trend verschillen in levensverwachting tussen de geaffilieerde

Caribische staten en hun westerse bestuurders verband houden met hun mate van politieke

autonomie. We laten zien dat het verloop van de levensverwachting tussen 1980 en 2014 gunstiger

was voor Caribische gebieden met een sterke politieke band met hun westerse bestuurder,

Martinique, Guadeloupe en de Britse eilanden, dan voor Caribische staten met een hoge mate van

autonomie, de Amerikaanse Maagdeneilanden, Puerto Rico, Aruba en Curaçao. In vergelijking met

de westerse bestuurder zijn de onderliggende trends in doodsoorzaken vergelijkbaar tussen de

Caribische staten en weerspiegelen daarom mogelijk regiospecifieke factoren. De politiek

geaffilieerde Caribische staten verschillen echter sterk in de omvang van de sterfteveranderingen,


173

vooral voor hart- en vaatziekten en externe oorzaken. De minst gunstige trends worden gevonden

voor geaffilieerde gebieden met een hoge mate van politieke autonomie, Aruba en Curaçao.

In hoofdstuk 6 van dit proefschrift bekeken we of, en waarom, 16 Caribische staten verschillen

in hun prestaties voor gezondheidsbeleid tussen 2010 en 2015, op 11 verschillende beleidsterreinen

die verband houden met regionale gezondheidsdoelstellingen van de Pan-Amerikaanse

Gezondheidsorganisatie (PAHO): 1. HIV/AIDS, 2. Infectieziekten, 3. Kanker screening, 4. Tabak,

5. Vruchtbaarheid, zwangerschap en bevalling, 6. Gezondheid van kinderen, 7. Diabetes, 8.

Hypertensie, 9. Alcohol, 10. Verkeersveiligheid en 11. Geweld. Onze bevindingen suggereren dat

variaties in gezondheidsuitkomsten tussen Caribische staten gerelateerd zijn aan verschillen in de

implementatie van effectief gezondheidsbeleid. Kortom, Caribische staten met betere

gezondheidsuitkomsten voor de 11 beleidsterreinen in deze studie hadden vaker “best practice”

gezondheidsmaatregelen ingevoerd. Martinique, Cuba en Guadeloupe scoorden het hoogst voor

hun prestaties in gezondheidsbeleid, en Guyana, Belize en Suriname het laagst. Aruba had de

laagste prestatiescore van de vijf politiek geaffilieerde Caribische staten in dit onderzoek, wat laat

zien dat de Arubaanse prestaties op deze beleidsterreinen suboptimaal zijn. Wat de

gezondheidsresultaten betreft, presteert Aruba aanzienlijk ongunstiger voor sterfte aan

borstkanker, longkanker en verkeersongevallen, en voor tienerzwangerschappen. De meest recent

beschikbare sterftegegevens van Curaçao dateren uit 2007, waardoor we moesten afzien van een

vergelijkbare analyse voor Curaçao.

We vonden nogmaals een sterke associatie met politieke status: Caribische staten die politiek

verbonden zijn aan een westers land hadden hogere prestatiescores voor het gezondheidsbeleid.

De prestaties van het gezondheidsbeleid zijn ook beter in gebieden waar de bevolkingsdichtheid,

het bbp per hoofd van de bevolking en de controle op corruptie hoog zijn, en gebieden die minder

religieus gefragmenteerd zijn. In onze resultaten onderscheidt Cuba zich door haar politieke

onafhankelijkheid, haar lage economische ontwikkeling en haar uitzonderlijke

gezondheidsresultaten. De mogelijke sterftevermindering voor oorzaken die door “best practice”

gezondheidsbeleid kunnen worden vermeden verschilt sterk tussen Caribische staten en is

aanzienlijk voor het Caribisch gebied als geheel: als de sterftecijfers van Martinique in andere

Caribische staten zouden gelden, zou een algemene sterftevermindering van 12% kunnen worden

verkregen.

Samenvattend geeft dit proefschrift een beter inzicht in de gezondheidssituatie van de bevolking

in de Nederlands-Caribische eilanden, en de factoren die verband houden met deze


174

gezondheidssituatie, de politieke context en de prestaties van het gezondheidsbeleid in het

bijzonder. We laten zien dat de gezondheidstoestand van de Nederlands-Caribische bevolking op

de eilanden Aruba en Curaçao ongunstiger is dan in Nederland en dan in andere politiek

geaffilieerde Caribische staten. Zo sterven mensen op Aruba en Curaçao vaker dan in Nederland,

Guadeloupe en Martinique aan oorzaken die vermijdbaar zijn door tijdige en effectieve

gezondheidszorg, en/of interventies op het gebied van volksgezondheid en preventie: borstkanker,

baarmoederhalskanker, ischemische hartziekten, cerebrovasculaire aandoeningen, longontsteking,

nefrose/nefritis en verkeersongevallen. Op Curaçao leveren HIV/AIDS, dikke darm- en

endeldarmkanker, diabetes, hypertensie, perinatale sterfgevallen en geweld een belangrijke

bijkomende bijdrage aan de hogere sterfte onder de bevolking. Dit suggereert dat een belangrijke

oorzaak van de ongunstigere sterftecijfers in de Nederlands-Caribische eilanden is dat de lokale

overheden tot dusver niet optimaal hebben ge nvesteerd in de gezondheidsbehoeften van hun

bevolking. Het is daarom waarschijnlijk dat de gezondheidsuitkomsten in de Nederlands-

Caribische eilanden kunnen verbeteren met gerichte acties die de toegang en kwaliteit van de

gezondheidszorg verbeteren (hoofdstuk 3) en de uitvoering van “best practice”

beleidsmaatregelen stimuleren (hoofdstuk 6).

Evenals eerdere studies tonen we een positief effect aan van een sterke politieke band met een

westers land op de volksgezondheid in Caribische staten en, daarmee samenhangend, betere

prestaties in het gezondheidsbeleid (hoofdstuk 4, 5 en 6). De studies in dit proefschrift voegen

op twee belangrijke manieren aan de bestaande kennis toe. Ten eerste laten we zien dat de

levensverwachting in de huidige soevereine staten tijdens de koloniale periode al ongunstiger was

dan in de huidige geaffilieerde staten, dat dekolonisatie samen viel met een vermindering van de

groei in levensverwachting, en dat de kloof in levensverwachting tussen huidige sovereine en

geaffilieerde staten tot op de dag van vandaag blijft toenemen (hoofdstuk 4). Ten tweede tonen

we aan dat de verschillen van de trends in levensverwachting van de Caribische gebieden die

politiek verbonden zijn met Frankrijk, het VK en de VS en Nederland de wijze waarop deze

gebieden bestuurd worden weerspiegelen. De trends in levensverwachting zijn gunstiger in

gebieden met weinig politieke autonomie, Martinique, Guadeloupe, en de Britse eilanden, dan in

gebieden met meer politieke autonomie, Puerto Rico, de Amerikaanse Maagdeneilanden, Aruba

en Curaçao (hoofdstuk 5). De mate van politieke autonomie vertaalt zich in verschillende

bestuursstructuren op het gebied van de gezondheid, wat suggereert dat lokale overheden in

Caribische gebieden waar gezondheids(zorg)organisaties samenwerken met hun overeenkomstige

organisaties in het land van de westerse bestuurder succesvoller zijn in het verbeteren en

beschermen van de gezondheid van hun bevolking. Martinique en Guadeloupe hadden


175

bijvoorbeeld vaker “best practice” gezondheidsmaatregelen geimplementeerd (hoofdstuk 6),

waarschijnlijk doordat gebieden met minder politieke autonomie meer menselijk kapitaal,

financiële middelen, investeringen en technologieën tot hun beschikking hebben om

gezondheidsproblemen te identificeren en aan te pakken, maar ook doordat zij effectievere

bureaucratieën hebben om plannen in uitvoering te brengen.

Omgekeerd geeft dit ook aan dat de halfslachtige politieke situatie tussen de Nederlands-Caribische

eilanden en Nederland, waarin discussies over politieke autonomie en financiële beheersbaarheid

van de eilanden in de afgelopen deccenia overheersten, heeft bijgedragen aan de huidige

gezondheidssituatie in de Nederlandse Cariben. Hoewel dit kan betekenen dat besprekingen over

het constitutionele karakter van het Koninkrijk der Nederlanden en de bevoegdheden van de

Rijksministerraad nodig zijn, impliceren onze resultaten vooral dat een groter bewustzijn van de

toenemende gezondheidsachterstand ten opzichte van Nederland en van andere politiek

geaffilieerde Caribische staten nodig is, alsook een vastberadenheid om deze

gezondheidsachterstanden aan te pakken.

Toekomstig gezondheidsonderzoek op de Nederlands-Caribische eilanden zou zich meer moeten

richten op het monitoren van de volksgezondheid en de gezondheidszorg om besluitnemers te

voorzien van relevante en tijdige informatie. Voor het terugdringen van de

gezondheidsachterstanden raden we aan om prioriteit te geven aan beleidsterreinen waarin

suboptimaal wordt gepresteerd, omdat effectieve interventies op deze gebieden waarschijnlijk de

grootste gezondheidswinst opleveren onder de bevolking. Als onze bevindingen over de gunstigere

gezondheidssituatie op Martinique, Guadeloupe en de Britse eilanden enige inspiratie bieden, dan

is het wel dat de haalbaarheid van een succesvolle uitvoering van deze aanbevelingen profiteert

van samenwerkingen met Nederlandse (gezondheids)organisaties.

Resúmen na papiamentu


178

Hulanda ta kana dilanti riba tereno di monitoreo di salú di poblashon, pero tokante poblashon di

e teritorionan di Reino Hulandes den Karibe no tin muchu informashon di salú disponibel. Kisas

esei tin di aber ku loke Sir Michael Marmot a yega di bisa, esta ku “generalmente, kaminda hende

ta mas pober, informashon riba salú tambe ta mas pober”. Riba e dos islanan mas grandi di Karibe

Hulandes, Aruba i Kòrsou, ekspektativa di bida ta mas abou kompará ku na Hulanda, i mortalidat

di yu chikitu i mama ta mas haltu. Apénas ta konosí kiko ta e kousanan tras di e sifranan ménos

faborabel na Karibe Hulandes. E splikashon ku sa bin padilanti ta ku esaki tin di aber ku sierto

karakterístika regional, biológiko i kultural; ma kiko spesífikamente no ta ser menshoná. Pero e

deskripshonnan genériko akí no ta sufisiente pa por formulá strategia di maneho pa hisa nivel di

salú di e habitantenan di Karibe Hulandes.

Meta prinsipal di e tésis akí ta di duna un mihó bista di e situashon di salubridat na Karibe

Hulandes, i mustra kua ta e faktornan ku ta hunga un ròl den e situashon akí. Spesífikamente, nos

ta enfoká riba konteksto polítiko i prestashon di maneho di salubridat. Pa hasi esei, nos a usa dato

di enkuesta representativo di salú, registro di mortalidat, base di dato internashonal harmonisá i

rapòrt nashonal. En bista di e aspekto multidisiplinario di e faktornan ku ta determiná salú di un

poblashon, nos ta inkorporá informashon tumá for di literatura di un variedat di disiplina:

salubridat públiko, medisina, siensia polítiko, siensia organisashonal, ekonomia i sosiologia.

Promé, nos a investigá kon ta pará ku salubridat di poblashon di Karibe Hulandes kompará ku

poblashon di Hulanda i poblashon di e otro estadonan den Karibe. Nos a traha ku tres diferente

manera di midi salú: loke e doño di kurpa mes ta ekspresá tokante su salú, ekspektativa di bida, i

sifra di mortalidat ku kousa spesífiko. Nos a konstatá ku salú di poblashon di Karibe Hulandes ta

ménos bon ku di Hulanda (kapítulo 2 i 3) i ménos bon ku di e estadonan karibense ku ta afiliá

polítikamente na Fransia, Reino Uní (UK) i Merka (US) (kapítulo 5).

Na kapítulo 2, nos a hasi un komparashon ‘tripartit’ entre poblashon di Karibe Hulandes bibá na

Kòrsou, inmigrante di Karibe Hulandes bibá na Hulanda, i persona no inmigrante bibá na Hulanda.

E resultadonan di kapítulo 2 ta indiká, kisas inesperadamente, ku e poblashon di Karibe Hulandes

na Kòrsou ta bisa ku nan tin mihó salú físiko i mental kompará ku e inmigrantenan di Karibe

Hulandes bibá na Hulanda. Salú di e inmigrantenan di Karibe Hulandes bibá na Hulanda tabata

ménos bon ku di persona no inmigrante na Hulanda tambe. E resultadonan akí a persistí asta

despues ku nos a hasi korekshon pa tuma na kuenta sierto variashon sosial i demográfiko (manera

edat, estado sivil, nivel di enseñansa i situashon di empleo), faktor di riesgo (uso di tabako i alkohòl,

sobrepeso i aktividat físiko) i nivel di urbanisashon (na Hulanda). Pa loke ta sobrepeso/obesidat i


179

diabétis, e situashon ta otro. Serka e poblashon di Karibe Hulandes bibá na Kòrsou, prevalensia di

sobrepeso/obesidat (hende hòmber so) i diabétis (tur dos sekso) ta similar ku serka e

inmigrantenan bibá na Hulanda, i e ta mas haltu ku serka e personanan no inmigrante na Hulanda.

Pa Aruba no tabatin dato ku por a usa pa hasi un komparashon den e investigashon akí.

Na kapítulo 3, e investigashon ta mustra ku, di 1988 pa 2014, e nivel di morto prevenibel (esta,

morto debí na kousa ku lo no mester hiba na morto si e persona haña kuido di salú efektivo na

tempu) tabata konsistentemente mas haltu na Aruba i na Antia/Kòrsou, kompará ku Hulanda. Si

e sifra di mortalidat prevenibel na Aruba baha te na mas o ménos e nivel di Hulanda, ekspektativa

di bida di hende hòmber i muhé na Aruba lo subi ku 1,19 aña i 0,72 aña respektivamente. Na

Kòrsou, e oumento lo ta di 2,06 aña i 2,33 aña. Si e sifra di mortalidat debí na enfermedat

sirkulatorio prevenibel (manera enfermedat iskémiko di kurason, enfermedat serebrobaskular i

enfermedat relashoná ku preshon haltu) baha te na mas o ménos e nivel di Hulanda, ekspektativa

di bida di hende hòmber i muhé na Aruba lo subi ku 1,02 aña i 0,56 aña respektivamente. Na

Antia/Kòrsou, e lo subi ku 0,65 i 0,55 aña. Morto prevenibel ta kontribuí supstansialmente na e

sifra di mortalidat di Antia/Kòrsou. Morto prevenibel ta responsabel pa 0,79 aña (hòmber) i 0,90

aña (muhé) di e diskrepansia total den mortalidat ku tin entre e islanan akí i Hulanda.

Nefrítis/nefrósis tambe ta un faktor importante na Antia/Kòrsou. E ta responsabel pa 0,12 aña di

e diskrepansia total den sifra di mortalidat serka hende hòmber i 0,21 aña serka hende muhé.

Ademas, si mortalidat debí na kanser na pechu baha te na e nivel di Hulanda, ekspektativa di bida

di hende muhé na Antia/Kòrsou lo subi ku 0,22 aña i na Aruba ku 0,21 aña.

E resultadonan presentá na kapítulo 5 ta mustra ku, di 1980 pa 2014, ekspektativa di bida di hende

hòmber na Aruba i Antia/Kòrsou a baha i di hende muhé a stagna, i e desaroyo akí a pone ku e

diskrepansia den ekspektativa di bida entre e region akí i Hulanda a sigui krese atraves di aña. Na

mes momento, atraves di aña, ekspektativa di bida den e teritorionan karibense ku tin relativamente

poko outonomia, manera Martinique, Guadeloupe i e islanan britániko a kana pareu ku esun di e

pais ku ta goberná nan. Un análisis mas detayá di e diferensianan entre ekspektativa di bida na e

paisnan di Oksidente i na nan teritorionan afiliá ta mustra ku enfermedat kardiobaskular i faktor

eksterno ta kontribuí supstansialmente na e sifranan di mortalidat mas haltu di e teritorionan den

Karibe. Sin embargo, e diskrepansia den mortalidat entre e diferente paisnan karibense i nan

paisnan oksidental ta varia bastante. Un diferensia entre e teritorionan hulandes i e otro partinan

di Karibe ta ku na e teritorionan hulandes no tabatin un mehora supstansial den e periodo di 1980

pa 2014 riba tereno di mortalidat debí na enfermedat kardiobaskular. Ademas, e kantidat di

homisidio na Antia/Kòrsou a oumentá drástikamente den e periodo akí, meskos ku a sosodé


180

tambe den e teritorionan estadounidense: Puerto Rico i U.S. Virgin Islands. Aksidente den

transporte tambe ta un di e faktornan prinsipal ku a subi e sifra di mortalidat ku kousa eksterno na

Antia/Kòrsou; na Aruba esaki ta e faktor prinsipal.

Di dos, nos a studia dos faktor ku por tin influensia riba e variashonnan den salú di poblashon bou

di e diferente estadonan karibense, esta, konteksto polítiko i prestashon di maneho di salubridat.

Nos a para ketu na tres aspekto di konteksto polítiko: státus polítiko aktual, dekolonisashon, i nivel

di outonomia polítiko. Generalmente, loke nos a mira ta ku dekolonisashon a oumentá e

diskrepansia di ekspektativa di bida entre e paisnan di Karibe ku awor ta soberano i esunnan ku ta

afiliá na un pais di Oksidente. Tambe nos a konstatá ku na e paisnan di Karibe ku tin un afiliashon

fuerte ku un pais oksidental, generalmente salú di e poblashon i prestashon di maneho di salubridat

ta na un mihó nivel.

Na kapítulo 4, nos ta studia kon e diskrepansia den ekspektativa di bida entre e paisnan di Karibe

ku aktualmente ta soberano i esunnan ku ta afiliá a desaroyá atraves di tempu. Loke nos a mira ta

ku for di promé ku e proseso di dekolonisashon a tuma lugá meimei di siglo 20, ekspektativa di

bida na e paisnan ku aktualmente ta soberano ya kaba tabata mas abou ku na e paisnan ku

aktualmente ta afiliá. Dekolonisashon, esta, e proseso di independisashon polítiko di un kolonia, a

bin kompañá pa un temporada, den e dékada despues di independensia, di ménos oumento den

ekspektativa di bida, i e kambio akí no tabata konektá ku otro kambio similar den prestashon

ekonómiko. Den e dékada di 1990, despues di e periodo di dekolonisashon, e diskrepansia den

ekspektativa di bida entre e estadonan ku aktualmente ta soberano i esunnan afiliá a sigui krese

mas lihé, i e ta sigui krese te dia djawe.

Na kapítulo 5, nos ta wak si tin un relashon entre e grado di outonomia polítiko ku un estado

afiliá den Karibe tin i kon diferente su ekspektativa di bida ta kompará ku e pais oksidental ku e ta

afiliá na dje. Loke nos a mira ta ku di 1980 pa 2014, na e paisnan di Karibe ku tin un laso fuerte ku

un pais oksidental ku ta goberná nan, esta, Martinique, Guadeloupe i e islanan britániko,

ekspektativa di bida a desaroyá mas favorabel kompará ku e estadonan karibense ku tin un nivel

mas haltu di outonomia, esta, U.S. Virgin Islands, Puerto Rico, Aruba i Kòrsou. E kousanan di

morto na e diferente estadonan karibense a desaroyá na un manera paresido, si kompará nan ku

na e paisnan oksidental ku e estadonan akí ta afiliá na dje. Esei por indiká ku e kousanan di morto

den Karibe tin di aber ku sierto faktor ku ta spesífiko pa e region akí. Sin embargo, e manera ku

sifra di mortalidat a desaroyá atraves di tempu na e diferente estadonan karibense ku afiliashon

polítiko ta varia bastante, prinsipalmente pa loke ta trata enfermedat kardiobaskular i kousa


181

eksterno. Kaminda e desaroyo akí tabata ménos faborabel ta den e teritorionan afiliá ku tin un

nivel haltu di outonomia polítiko, esta, Aruba i Kòrsou.

Na kapítulo 6 di e tésis akí, nos ta investigá si akaso, i dikon, tin diferensia entre 16 estado

karibense den e periodo di 2010 pa 2015 pa loke ta prestashon di maneho di salubridat. Nos a

konsiderá 11 tereno di maneho ku ta relashoná ku e metanan regional di salubridat fihá pa

Organisashon Panamerikano di Salú (PAHO): 1. HIV/AIDS, 2. Enfermedat infeksioso, 3. Tèst

preventivo di kánser, 4. Tabako, 5. Fertilidat, embaraso i parto, 6. Salú di mucha, 7. Diabétis, 8.

Preshon haltu, 9. Alkohòl, 10. Siguridat den tráfiko i 11. Violensia. Nos resultadonan ta indiká ku

e diferensianan entre e diferente estadonan karibense riba tereno di salú por tin di aber ku diferensia

den implementashon di maneho di salubridat efektivo. Ku otro palabra, frekuentemente, ta e

estadonan karibense ku a introdusí medida di salubridat basá riba “best practice” ta skor mihó pa

loke ta e 11 terenonan di maneho den e estudio akí. Martinique, Cuba i Guadeloupe ta esunnan ku

a skor mas haltu pa prestashon di maneho di salubridat. Guyana, Belize i Sürinam ta e paisnan ku

a skor mas abou. Di e sinku estadonan karibense polítikamente afiliá den e investigashon akí, Aruba

a skor mas abou pa prestashon di maneho, loke ta indiká ku Aruba su prestashon di maneho no ta

optimal. Pa loke ta resultado di salú, Aruba su prestashon ta bastante ménos faborabel pa loke ta

morto debí na kánser na pechu, kánser na pulmon, morto den tráfiko i embaraso hubenil. Pa

Kòrsou no tin dato di mortalidat mas resien ku 2007 i debí na esei no por a hasi un análisis similar

pa Kòrsou.

Aki tambe nos por mira ku un estado su státus polítiko ta hunga un papel importante: E estadonan

karibense ku ta afiliá polítikamente na un pais oksidental a skor mas haltu pa loke ta prestashon di

maneho di salubridat. Ta asina ku, generalmente, bo ta topa mihó prestashon di maneho di

salubridat den área kaminda densidat di poblashon, produkto interno bruto pa kabes di poblashon

i kontrol di korupshon ta haltu i kaminda tin ménos fragmentashon religioso. Den nos

resultadonan, Cuba ta un eksepshon en bista di su independensia polítiko, su desaroyo ekonómiko

limitá i su resultadonan eksepshonal riba tereno di salú. Si wak kuantu mortalidat ku kousa

prevenibel por baha si apliká maneho di salubridat basá riba “best practices”, e kontesta ta varia

bastante di un estado di Karibe pa otro, pero pa Karibe den su totalidat, e potensial di redukshon

ta supstansial. Si e otro estadonan di Karibe tabatin e mesun sifra di mortalidat ku Martinique, esei

lo nifiká un redukshon general di 12% di mortalidat den Karibe.

Resumiendo, e tésis akí ta duna un mihó bista di kondishon di salú di poblashon di e islanan

hulandes den Karibe i tambe di e faktornan ku tin di aber ku nan kondishon di salú,


182

partikularmente konteksto polítiko i prestashon di maneho di salubridat. Nos ta mustra ku

kondishon di salú di poblashon na Aruba i Kòrsou ta ménos faborabel ku na Hulanda i algun otro

estado karibense ku ta polítikamente afiliá. Kompará ku Hulanda, Guadeloupe i Martinique por

ehèmpel, na Aruba i Kòrsou hende ta muri mas tantu debí na kousa ku por a prevení mediante

kuido di salú na tempu i efektivo i/òf mediante intervenshon riba tereno di salubridat i prevenshon.

Ta trata akí di kousa manera kánser na pechu, kánser na boka di matris, enfermedat iskémiko di

kurason, enfermedat serebrobaskular, pulmonia, nefrósis/nefrítis i aksidente di tráfiko. Na

Kòrsou, HIV/AIDS, kánser na tripa diki i na rekto, diabétis, enfermedat di kurason debí na

preshon haltu, morto perinatal i violensia ta kontribuí supstansialmente na e sifra di morto mas

haltu. Esei ta indiká ku e sifranan di mortalidat ménos faborabel na Aruba i Kòrsou por tin hopi

di aber ku e echo ku, te asina leu, e gobièrnunan lokal no a invertí na un manera optimal den e

nesesidatnan di salubridat di nan poblashon. P’esei, ku akshon dirigí pa mehorá akseso na kuido

di salú i kalidat di kuido (kapítulo 3) i si stimulá medida di maneho ku ta basá riba “best practices”

(kapítulo 6), probablemente e resultadonan di salubridat riba e islanan di Karibe Hulandes por

mehorá.

E investigashon akí, meskos ku algun otro investigashon anterior, ta mustra kon un laso fuerte ku

un pais oksidental ta kontribuí na un mihó nivel di salubridat i mihó prestashon di maneho di

salubridat den e estadonan karibense (kapítulo 4, 5 i 6). E investigashonnan den e tésis akí a aportá

konosementu nobo i relevante riba dos diferente área. Di promé, nos a mustra ku for di tempu

kolonial ya ekspektativa di bida den e estadonan ku aktualmente ta soberano tabata ménos

faborabel ku den e estadonan ku ta afiliá aktualmente, ku e proseso di dekolonisashon a bai

kompañá pa un periodo di ménos oumento di ekspektativa di bida, i ku te dia djawe e diskrepansia

den ekspektativa di bida entre e estadonan ku aktualmente ta soberano i esunnan ku aktualmente

ta afiliá ta sigui krese (kapítulo 4). Di dos, nos a mustra ku e diskrepansianan den e manera ku

ekspektativa di bida a desaroyá den e teritorionan karibense ku ta polítikamente afiliá na Fransia,

Reino Uní, Merka i Hulanda ta un refleho di e manera ku e teritorionan akí ta ser goberná. Den e

teritorionan ku poko outonomia polítiko, esta Martinique, Guadeloupe i e islanan britániko,

ekspektativa di bida a desaroyá mas faborabel kompará ku den e teritorionan ku tin mas outonomia

polítiko: Puerto Rico, U.S. Virgin Islands, Aruba i Kòrsou (kapítulo 5). E midí di outonomia

polítiko ta keda reflehá den e diferente strukturanan di gobernashon riba tereno di salubridat, i

esaki por ta mustra riba e echo ku, den e teritorionan karibense kaminda e organisashonnan di

salubridat ta kolaborá ku e organisashonnan den e pais oksidental ku nan ta afiliá na dje, e

gobièrnunan lokal tin mas éksito den mehorá i protehá salú di nan poblashon. Martinique i

Guadeloupe, por ehèmpel, a implementá mas tantu medida di salubridat basá riba “best practice”


183

(kapítulo 6). Esei probablemente ta debí ku e teritorionan ku ménos outonomia polítiko tin mas

kapital humano, fondo, invershon i teknologia na nan disposishon pa por identifiká i atendé

problema di salubridat, i debí ku e burokrasia den e teritorionan ei ta mas efisiente i por implementá

plan mihó.

Di otro banda, e resultadonan akí ta mustra tambe riba e efekto di e situashon polítiko indesiso ku

ta eksistí entre e islanan di Karibe Hulandes i Hulanda. Pa algun dékada kaba, ta outonomia polítiko

i maneho finansiero di e islanan a dominá e diálogo entre e partinan di Reino. Tur esaki tambe a

kontribuí na e situashon di salubridat aktual na Karibe Hulandes. Aunke e resultadonan akí por

mustra riba un nesesidat pa deliberá riba e karakter konstitushonal di Reino Hulandes i e

responsabilidatnan ku Konseho di Minister di Reino tin, loke nos ke mustra na promé lugá ta ku

falta mas konsientisashon tokante e atraso kresiente riba tereno di salubridat den Karibe Hulandes

kompará ku Hulanda i tambe kompará ku otro estadonan karibense afiliá, i ku mester tin mas

determinashon pa atendé e atraso akí.

Den futuro, investigashon riba tereno di salú riba e islanan di Karibe Hulandes lo mester konsentrá

mas riba monitoreo di salubridat di poblashon i monitoreo di kuido di salú, pa asina esnan ku tin

ku tuma desishon por tin informashon relevante i na tempu. Pa redusí e atraso riba tereno di

salubridat, nos rekomendashon ta pa duna prioridat na e areanan di maneho kaminda e prestashon

no ta optimal, ya ku ta riba e terenonan akí probablemente un intervenshon efektivo lo resultá den

mas benefisio pa salú di e poblashon. Si loke nos a konstatá enkuanto e mihó situashon di

salubridat na Martinique, Guadeloupe i e islanan britániko por sirbi komo inspirashon, loke mester

saka for di esei ta ku un kolaborashon mas estrecho ku e organisashonnan (di salubridat) hulandes

lo oumentá e posibilidat pa e rekomendashonnan akí por keda implementá eksitosamente.

Acknowledgements

Acknowledgements

186

This thesis once began as a big, messy research proposal that listed rather vague and impracticable

ideas on how to explore the situation of population health in the Dutch Caribbean. It required

more than just my own hands and thinking power to chisel these ideas to something worthy of a

PhD-degree for which, first and foremost, I have my two supervisors to thank.

Johan, your drive and sharp mind is inspiring and admirable, and our discussions have been some

of the most rewarding experiences of my PhD. Thank you for your insights, and for your trust to

take me on board as an external PhD-candidate. Hans, I am forever captivated by your quest to

make scientific insights serve the public good, blurring the line between the two worlds of slow-

paced academics and fast-paced government. Your drive for policy relevance is refreshing and

stimulating. Thank you for your support, and for giving me the space to pursue my research

interests.

I have been fortunate enough to have had many mentors during my career so far, who each have

stimulated me to aim high and to stay endlessly curious in their own way: Theo Harmsen, John

Rossen, Marika Nosten-Bertrand, Bruno Giros, Niels Galjart, Frank Grosvelt, Jaap Geenen,

Ronald Westerhof, Jan Huurman and Prof. Bob Pinedo.

My further thanks go to Gera Christina and Jeanine Constansia-Kook for their guidance and

support within the Ministry. To my fellow researchers Renske Pin and Merel Griffith-Lendering

for joining forces in research. To co-author Carolien van den Brink for her amazing ability to

clearly grasp the harmonization of differences between the datasets of the NGE and the Dutch

Health Monitor. To Mildred Albertoe for making sure I was well hydrated and could work in a

spotless office. To Jonathan Plantijn for the good times we shared at VIC. To Kevin Reina,

Shantilly Christina, and Charminy Rosario, for brightening up the VIC-office with your energy and

helping us with our efforts.

To my grandparents Piet and Naan, who have taught me to meet life’s challenges with grace and

persistence. Thank you for your unwavering believe in me. You have given me my roots. To my

mother and father, who got me used to thinking, and acting, outside-the-box. To my stepfather

Stefan, mother-in-law Bianca, and sister-in-law Michelanca, for always lending us a helping hand.

To my partner-in-crime Raygen: you are awesome. What a ride it has been, and your support was

indispensable. And to our daughters, Jazmine and Solange: the two of you are the foundation of

everything good in my life.

And lastly, to Curaçao. The island that feels like home and that I love. Warts and all.

About the author

About the author

188

Soraya Verstraeten was born in Breda, the Netherlands (1981). While working as an analyst

in clinical microbiology at the University Medical Center of Utrecht, she simultaneously studied

for her BAS in molecular biology, which she completed cum laude. During her time in Utrecht,

she connected with people who stimulated her drive to explore how the world works and

became interested in health research - specifically, in finding solutions for real-world

problems using scientific methods. This led to her pursuit of a Msc degree in biomedical

sciences at Utrecht University, which she completed in 2008. In the same year, she was

nominated for the TopTalent scholarship programme of the NWO and started a PhD at the

department of Cell Biology at the Erasmus MC. After a year and a half, Soraya left her PhD

unfinished, as the disconnect between cell biology and society proved a stumbling block for her

motivation. Next to microbiology and cell biology, she had the good fortune of being able to

pursue her academic curiosities in the fields of (psycho)pharmacology, neurobiology, behavioral

sciences, cardiology and molecular biology, before settling down in the field of public health

In 2010, Soraya moved to the beautiful island of Curaçao. There, she started working at the local

Government in the middle of its historic constitutional change towards an autonomous country

within the Kingdom of the Netherlands. In 2012, she co-founded the Institute for Public Health

in Curaçao (Volksgezondheid Instituut Curaçao, VIC), where she coordinates data collection on

health and healthcare through population-based surveys and data registration systems to support

evidence-based policy making. As the principal investigator, she has led efforts on several large

health studies, among which the national health surveys in 2013 and 2017 and the Global School-

based Student Health Survey (GSHS) in 2015.

Wanting to know how to use sound research to promote good population health in a small island

setting, Soraya went on to pursue her PhD with Prof. dr. Johan Mackenbach and Prof. dr. Hans

van Oers. She completed her PhD at the Institute for Public Health (Ministry of Health,

Environment and Nature, Curaçao) and the Department of Public Health (Erasmus MC,

Rotterdam) in April 2020.

In her work, Soraya is driven by bridging the gap between scientific ideas and their pragmatic

application for real-world problems. Next to research, she teaches general and organic Chemistry

at the St. Martinus University School of Medicine and is involved with several initiatives that aim

to develop and implement solutions to social challenges in Curaçao.

List of publications

List of publications

190

This thesis

Verstraeten SPA, van Oers, HAM, & Mackenbach, JP. (2019) Health Policy Performance in 16 Caribbean States, 2010–2015. American Journal of Public Health, 109, 4, 626-632. https://doi.org/10.2105/AJPH.2018.304733

Verstraeten SPA, van Oers, HAM, & Mackenbach, JP. (2019). Differences in life expectancy between four Western countries and their Caribbean dependencies, 1980-2014. European Journal of Public Health. Ckz102. https://doi.org/10.1093/eurpub/ckz102.

Verstraeten SPA, van Oers, HAM, & Mackenbach, JP. (2019). Contribution of amenable mortality to life expectancy differences between the constituent countries of the Kingdom of the Netherlands. Submitted.

Verstraeten SPA, van den Brink CL, Mackenbach JP, & van Oers, HAM (2018). The health of Antillean migrants in the Netherlands: a comparison with the health of non-migrants in both the countries of origin and destination. International Health. 10, 4, 258-267. https://doi.org/10.1093/inthealth/ihy026

Verstraeten SPA, van Oers, HAM, & Mackenbach, JP. (2016). Decolonization and life expectancy in the Caribbean. Social Science & Medicine, 170, 87–96. https://doi.org/10.1016/j.socscimed.2016.08.048

Other (selection)

Verstraeten SPA. Health indicators in Curaçao 2018. Willemstad, 2018.

Verstraeten SPA. Resultaten monitor wachttijden medisch specialistische zorg. 2017.

Verstraeten SPA., Griffith-Lendering M, Pin R. De Nationale Gezondheidsenquête 2017. Resultaten, methode en tabellen. Willemstad, 2017.

Pin R.R., Verstraeten SPA., Griffith-Lendering M. De Nationale Gezondheidsenquête 2017. Themarapport Arme wijken. Willemstad, 2017.

Verstraeten SPA. The Curaçao Global School-based Student Health Survey (GSHS) Study 2015. Willemstad 2016.

Verstraeten, SPA., Jansen, I. De Nationale Gezondheidsenquête 2013. Themarapport ouderen. Willemstad, 2013.

Verstraeten, SPA., Jansen, I., Pin, R.R., Brouwer, W. De Nationale Gezondheidsenquête 2013. Methodologie en belangrijkste resultaten. Willemstad, 2013.

PhD portfolio

PhD portfolio

192

PhD training, teaching and other activities

Candidate Soraya Petronella Adriana Verstraeten

Affiliation Institute for Public Health (VIC), Ministry of Health, Environment and Nature, Curaçao and Department of Public Health, Erasmus MC, Rotterdam

PhD period January 2015 – September 2019

Training

Courses – Public Health Year Workload (ECTS*)

Erasmus MC ESP11 Methods of Public Health Research 2015 0.7 ESP42 Methods of Health Services Research 2015 0.7 ESP45 Primary and Secondary prevention research 2015 0.7 ESP61 Social epidemiology 2015 0.7 ESP 63 Master class: Advances in Genomics Research 2015 0.4 ESP64 Masterclass: Advances in Epidemiologic Analysis 2015 0.4 ESP66 Logistic Regression 2015 1.4 Massive Open Online Courses (MOOC) PH555x Improving global healthcare focusing on quality and safety (HarvardX) 2014 1.7

PH201x Health and Society (HarvardX) 2014 1.7 PH210x United States Health policy (HarvardX) 2014 1.9 SW25x Global health: case studies from a biosocial perspective (HarvardX) 2014 1.7

PH231 Readings in global health (HarvardX) 2016 1.4 System thinking in Public health (John Hopkins University online) 2017 1.4 Community change in Public Health (John Hopkins Universityonline) 2017 1.4

Courses – General Erasmus MC Scientific Integrity 2015 0.3 Massive Open Online Courses (MOOC) University teaching 101 (John Hopkins University) 2014 0.9 Writing in the Sciences (John Hopkins University) 2016 1.4 Total 18.8 *1 ECTS = 28 hours of workload

PhD portfolio

193

Presentations (selection) Description Institution(s)/event Year(s) Presentation “Global School-based Student Health Survey (GSHS) Curaçao 2015. Health behaviours and protective factors”

Ministerial management teams, policy makers, school boards, school directors and foundations

2015-2017

Workshop “Health Behaviour of adolescents in Curaçao”

Annual teachers seminar protestant school board (VPCO)

2017

Presentation results “National Health Survey Curaçao 2017”

Ministerial management teams, policy makers, and foundations

2017-2018

Panel discussion “It’s time to get real about food”

CoWorld 2017

Presentation “Poverty and health in Curaçao” Conference “Poverty in Curaçao”, Dutch Caribbean Economists

2018

Presentation “A comparative analysis of population health in the Caribbean”

Annual symposium Public Health, Avalon University School of Medicine

2019

Presentation “Communicable diseases in Curacao, a comparison with the Netherlands and Aruba”

Parliament of Curaçao 2019

Teaching Course Institution Year(s) Training survey administrators, Global School-based Student Health Survey (GSHS) 2015

Institute for Public Health/Central Bureau of Statistics

2015

Training of survey administrators, National Health Survey Curaçao 2017

Institute for Public Health/Central Bureau of Statistics

2017

General and Organic Chemistry St. Martinus University, Faculty of Medicine

2012-now

Other activities Description Institution Year(s) Researcher/project manager Institute for Public Health 2012-nowReviewer Social Science & Medicine, Global Public

Health and International Journal of Health Policy and Management

2016-2019

Guidance committee “Exploration monitor juvenile criminal law Dutch Caribbean”

WODC, Ministry of Justice and Security, the Netherlands

2019

Population Health in the Dutch Caribbean

Documents

Transcript of Population Health in the Dutch Caribbean